CN113094529B - Image data processing method and device, electronic equipment and storage medium - Google Patents
Image data processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113094529B CN113094529B CN201911338762.1A CN201911338762A CN113094529B CN 113094529 B CN113094529 B CN 113094529B CN 201911338762 A CN201911338762 A CN 201911338762A CN 113094529 B CN113094529 B CN 113094529B
- Authority
- CN
- China
- Prior art keywords
- data
- initial data
- initial
- video memory
- image processor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/53—Querying
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/55—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of data processing, and provides a processing method and device of image data, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring initial data, and storing the initial data in a hard disk in a classified manner, wherein the initial data comprises a data type and a data volume; the method comprises the steps of transferring initial data which are the same in data type and have data quantity reaching a preset data quantity threshold value in a hard disk into a video memory of an image processor for classified storage; acquiring data to be retrieved, and judging whether the initial data with the same data type as the data to be retrieved is stored in a video memory of the image processor; if so, comparing the similarity between the data to be searched and the initial data in the video memory of the image processor, and taking the initial data with the similarity meeting the preset condition as target search data. The embodiment of the invention ensures the retrieval response time and simultaneously reduces the hardware cost.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a method and apparatus for processing image data, an electronic device, and a storage medium.
Background
The traditional image retrieval is based on CPU hardware resources, and the similarity calculation is carried out on the characteristic values of the image and the characteristic values of the database data, so that the CPU is not good at carrying out the operation with high parallelism due to the large operation amount during the characteristic value calculation, and the problem that the retrieval speed is slow can be obvious because the efficient and rapid retrieval cannot be carried out during the massive data retrieval. Along with the development of technology, aiming at the situations of huge data volume and more data types in an application system, the increase of the data volume occupies a large amount of storage space, and the retrieval in a CPU is high in cost and takes a large amount of time. It is apparent that the conventional technology has problems of slow response speed and high hardware cost when retrieving image data.
Disclosure of Invention
The embodiment of the invention provides a processing method of image data, which can accelerate response speed and reduce hardware cost.
In a first aspect, an embodiment of the present invention provides a method for processing image data, including:
acquiring initial data, wherein the initial data comprises a data type and a data volume;
classifying and storing the initial data in a hard disk according to the data type;
the method comprises the steps of transferring initial data which are the same in data type and have data quantity reaching a preset data quantity threshold value in a hard disk to a video memory of an image processor for classified storage;
Obtaining data to be searched, judging a storage area where the data type of the data to be searched is located, comparing the similarity in the storage area where the data type of the data to be searched is located, and taking initial data with the similarity meeting a preset condition as target search data.
In a second aspect, an embodiment of the present invention further provides a processing apparatus for image data, including:
the acquisition module is used for acquiring initial data, wherein the initial data comprises a data type and a data volume;
the storage module is used for classifying and storing the initial data in the hard disk according to the data type;
the transfer module is used for transferring the initial data which are of the same data type and have the data quantity reaching the preset data quantity threshold value in the hard disk to the video memory of the image processor for classified storage;
the first judging module is used for acquiring the data to be searched, judging a storage area where the data type of the data to be searched is located, comparing the similarity in the storage area where the data type of the data to be searched is located, and taking initial data with the similarity meeting a preset condition as target search data.
In a third aspect, an embodiment of the present invention further provides an electronic device, including: the image processing device comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to realize the steps in the image data processing method provided by the embodiment of the invention.
In a fourth aspect, an embodiment of the present invention further provides a computer readable storage medium, where a computer program is stored, where the computer program when executed by a processor implements steps in a method for processing image data provided by the embodiment of the present invention.
In the embodiment of the invention, initial data is acquired, the initial data is stored in a hard disk in a classified manner, and the initial data comprises data types and data amounts; the method comprises the steps of transferring initial data which are the same in data type and have data quantity reaching a preset data quantity threshold value in a hard disk into a video memory of an image processor for classified storage; acquiring data to be searched, and judging whether initial data with the same data type as the data to be searched is stored in a video memory of an image processor or not; if the data to be searched and the initial data are compared in the video memory of the image processor, and the initial data with the similarity meeting the preset condition is used as target search data. The invention stores the initial data in the hard disk in advance, then judges whether the data quantity of the initial data with the same data type reaches the preset data quantity threshold value, and transfers the reached initial data to the video memory of the image processor, and the initial data which is not reached is still stored in the hard disk; meanwhile, when the initial data is transferred to the video memory of the image processor, the data type of the initial data is classified and stored, so that the initial data corresponding to the data to be searched can be conveniently searched, all the initial data do not need to be traversed, and the searching response time is shortened.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for processing image data according to an embodiment of the present invention;
FIG. 2 is a flowchart of another method for processing image data according to an embodiment of the present invention;
FIG. 3 is a flowchart of another method for processing image data according to an embodiment of the present invention;
FIG. 4 is a flowchart of another method for processing image data according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an apparatus for processing image data according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of another image data processing apparatus according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of another image data processing apparatus according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of another image data processing apparatus according to an embodiment of the present invention;
Fig. 9 is a schematic structural diagram of another image data processing apparatus according to an embodiment of the present invention;
fig. 10 is a schematic structural diagram of another image data processing apparatus according to an embodiment of the present invention;
fig. 11 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, fig. 1 is a flowchart of a method for processing image data according to an embodiment of the present invention, which specifically includes the following steps:
s101, acquiring initial data, and storing the initial data in a hard disk in a classified manner, wherein the initial data comprises data types and data amounts.
In this embodiment, the electronic device on which the image data processing method operates may acquire the initial data through a wired connection manner or a wireless connection manner, or the like. It should be noted that the Wireless connection may include, but is not limited to, a 3G/4G connection, a WiFi (Wireless-Fidelity) connection, a bluetooth connection, a WiMAX (Worldwide Interoperability for Microwave Access) connection, a Zigbee (low power lan protocol, also known as the purple peak protocol) connection, a UWB (ultra wideband) connection, and other now known or later developed Wireless connection methods.
It should be noted that the electronic device may be various electronic devices that have a display screen and support web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.
Before massive data retrieval calculation is performed, initial data needs to be acquired as base data. The initial data may be automatically input data or manually input data. The initial data may be raw image data, and the data types of the initial data may be preset, and the data types may be used to classify the initial data, so as to avoid that all the initial data of the data types are stored in the same storage area. The data amount described above may represent the data size of the initial data, for example: the data amount was 20 ten thousand. The hard disk may be a CPU, which is one of the main devices of a computer, the function of which is mainly to interpret computer instructions and process data in computer software. The hard disk may be used to store the initial data described above.
Based on the difference of the data types, the initial data of the same data type can be subjected to class aggregation, and scattered initial data can be aggregated into one or more large classes. A plurality of memory areas may be provided in the hard disk, and the classified initial data may be correspondingly stored in the respective memory areas. The storage mode can be that the storage is sequentially classified according to the data types of the input initial data, when all storage areas in the hard disk store the initial data, the storage areas which have the same data types and are not full of the initial data can be sequentially searched from the first storage area, and the initial data which are input later are stored, for example: the initial data of four data types A, B, C, D are sequentially recorded and can be sequentially stored in four storage areas of the hard disk a, B, c, d, if 4 storage areas are shared in the hard disk, when the 4 storage areas all store the initial data of the corresponding data types, if the data type of the initial data stored currently is B, the B storage area can be sequentially searched, and the B initial data can be continuously stored in the B storage area. The search may also be a random search or an inverse search from the last memory area.
If the same storage area is occupied by the initial data with the same data type, other storage areas in the hard disk can be searched when the initial data is subsequently recorded, for example, the data type of the initial data which is currently recorded is B, and the storage area B is already full, the rest storage areas can be searched in sequence, and if the storage area c has enough storage areas, the data type of the initial data is B and is stored in the storage area c. Of course, the storage modes described above are only some of the storage modes in the embodiments of the present invention, and the storage modes are not limited in the present invention.
S102, the initial data which are the same in data type and have the data quantity reaching a preset data quantity threshold value in the hard disk are transferred to a video memory of the image processor for classified storage.
The image processor is a GPU, also called a display core, a vision processor, and a display chip, and is a microprocessor specially used for image operation on a personal computer, a workstation, a game machine, and some mobile devices (such as a tablet computer, a smart phone, etc.), and has high performance on processing logic operation. The above-mentioned video memory may also be referred to as a display memory, and the video memory may have capacities of different sizes, for example: the memory capacity of the memory includes 128MB, 256MB, 512MB, 1024MB,64MB, 128MB, 1TB, etc. The larger the data that needs to be buffered, the larger the occupied capacity will be.
The above-mentioned preset data amount threshold may be set according to the size of the data amount to be processed, for example: 1 hundred million initial data, require its response time to finish the search to all data within 3s, when just begin to input the data, input the initial data into hard disk and store, and the time of carrying on the violent search to every initial data is within 10us, search 10 ten thousand initial data take 1s, so regard 3s as a goal response time, can set up the above-mentioned preset data quantity threshold value as 25 ten thousand data. The above-mentioned preset data amount threshold is set for the initial data of any one data type.
The above classification of the initial data based on the data type can reduce the search and calculation range. Therefore, when the data is transferred to the GPU video memory, the data can be stored according to the data types of the data, and a plurality of storage areas are arranged in the GPU video memory, so that initial data which are the same in data types and have the data quantity exceeding a preset data quantity threshold value can be sequentially stored according to the sequence of the storage areas. Of course, the initial data whose data amount exceeds the preset data amount threshold may be randomly stored in each storage area, which is not limited in the embodiment of the present invention. And the data types are the same and the data quantity exceeds the preset data quantity threshold value, and the data are transferred to the GPU video memory for storage, so that the storage space of the CPU can be released, and the response speed of data retrieval is ensured.
S103, acquiring data to be searched, and judging whether initial data with the same data type as the data to be searched is stored in a video memory of the image processor.
The data to be retrieved may be data contained in a retrieval request sent by the user through the mobile terminal. The method comprises the steps of judging the data type of the data to be searched, searching whether the initial data of the data type is stored in the video memory of the GPU, if so, performing subsequent searching in the GPU, otherwise, storing the initial data of the same data type as the data to be searched in the CPU, and otherwise, performing subsequent searching in the GPU.
And S104, if the data to be searched and the initial data are compared in the video memory of the image processor, and the initial data with the similarity meeting the preset condition are used as target search data.
If the initial data with the same data type as the data to be searched exists in the video memory of the GPU, the initial data with the same data type corresponding to the data to be searched can be searched in the video memory, and the similarity comparison is carried out on the data to be searched and the initial data with the same data type, so that all the initial data can be searched one by one, and the problems of large data searching amount and long response time can be avoided.
The similarity may represent a degree of similarity between the initial data and the data to be retrieved, and the similarity comparison may represent calculation of a similarity between all the initial data and the data to be retrieved of the same data type. The preset condition may be that, based on all the calculated similarities, the topN similarities with the top ranking are selected, and initial data corresponding to the topN similarities with the top ranking may be used as the target search data, and the target search data is returned to the mobile terminal used by the user. Of course, the above-mentioned preset condition may also be a preset similarity threshold, and after all the similarities are calculated, initial data corresponding to the similarities reaching the preset similarity threshold may be used as target retrieval data, and the target retrieval data may be returned to the mobile terminal used by the user.
The mobile terminal may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, electronic book readers, players, laptop and desktop computers, and the like.
In the embodiment of the invention, initial data is acquired, wherein the initial data comprises a data type and a data volume; classifying and storing the initial data in a hard disk according to the data type; the method comprises the steps of transferring initial data which are the same in data type and have data quantity reaching a preset data quantity threshold value in a hard disk to a video memory of an image processor for classified storage; and obtaining the data to be searched, judging a storage area where the data type of the data to be searched is located, and comparing the similarity in the storage area to obtain the target search data. The invention stores the initial data in the hard disk in advance, then judges whether the data quantity of the initial data with the same data type reaches the preset data quantity threshold value, and transfers the reached initial data to the video memory of the image processor, and the initial data which is not reached is still stored in the hard disk; meanwhile, when the initial data is transferred to the video memory of the image processor, the data type of the initial data is classified and stored, so that the initial data corresponding to the data to be searched can be conveniently searched, all the initial data do not need to be traversed, and the searching response time is shortened. .
As shown in fig. 2, fig. 2 is a flowchart of another image data processing method provided in an embodiment of the present invention, which specifically includes the following steps:
s201, acquiring initial data, and identifying the data type of the initial data.
The data types described above may include data objects and data attributes. The data object and the data attribute may be basic parameters that constitute a data type, and the data object may be used as a first dimension, and the data attribute may be used as a second dimension. The data objects may include, but are not limited to, human faces (five sense organs, masks, sunglasses, hats), human bodies (body types, clothes, heights, weights, patterns on clothes, styles of clothes, bags on the back, things carried, etc.), static libraries, vehicle information (including user information, vehicle registration information, etc.), wearing, behaviors, etc., wherein the static libraries may refer to libraries storing some common codes that need to be repeatedly used. The data attributes may include, but are not limited to, information of blacklisted people, blacklisted vehicles, pursuers, etc. The specific data type of the initial data can be judged by acquiring the data object and the data attribute of the initial data.
After the data object and the data attribute of the initial data are acquired, the data object and the data attribute may be combined to obtain different data types, for example: the face of the blacklisted person, the wearing of the blacklisted person, the running action of the blacklisted person, the vehicle information of the blacklisted vehicle, and the like. The data types obtained by combining the data objects and the data attributes can be more than 15 types.
The data volume of the data attributes described above may vary depending on the data volume of the data object, for example: the data attribute is a blacklist person, the data object is a face, the corresponding data amount is 1 ten thousand, and if the data object is a human body, the corresponding data amount is 2 ten thousand. And may also vary according to different application scenarios, for example: the data attribute is blacklist vehicles, the data object is vehicle information, the data amount at the pedestrian street is 1000, and the data amount at the traffic road can be 10 ten thousand. The data object may be combined with a data attribute to form a data type corresponding to a plurality of initial data, for example: the number of initial data corresponding to the faces of the blacklist personnel is 10 ten thousand, and the number of initial data of running actions of the blacklist personnel is 2 ten thousand.
S202, merging initial data with the same data type, and storing the initial data with the same data type in the same hard disk, wherein the hard disk comprises a plurality of hard disk storage areas, and each hard disk storage area stores the initial data with one data type.
After the data types of all the initial data are identified, the initial data of the same data type can be combined, and the initial data of the same data type are stored in the same hard disk storage area. The hard disk has a plurality of hard disk storage areas, and initial data of one data type can be stored in each hard disk storage area. Therefore, the classified storage of the initial data is realized, different types of initial data can be classified and stored in the same hard disk, and when the data is searched, only the hard disk storage area where the initial data with the same data type as the data to be searched is needed to be searched, the search of the initial data with different data types is not needed, and the response speed and the search time are accelerated.
S203, the initial data which are the same in data type and have the data quantity reaching the preset data quantity threshold value in the hard disk are transferred to the video memory of the image processor for classified storage.
S204, acquiring data to be searched, and judging whether initial data with the same data type as the data to be searched is stored in a video memory of the image processor.
S205, if the data to be searched and the initial data are compared in the video memory of the image processor, and the initial data with the similarity meeting the preset condition are used as target search data.
In the embodiment of the invention, the data type of the initial data obtained by combining the data object and the data attribute is extracted by acquiring the initial data, and the initial data is stored in different hard disk storage areas in the same hard disk according to the data type. Therefore, the data types of the initial data are obtained by combining according to different dimensions, the corresponding storage is carried out based on the data types obtained after the combination, the storage space of the hard disk can be effectively and reasonably utilized, when the data to be searched is obtained and needs to be searched in the hard disk, the search can be carried out only in the initial data with the same data type as the data to be searched, the search of the initial data with different data types is not needed, and the response speed and the search time are accelerated.
As shown in fig. 3, fig. 3 is a flowchart of another image data processing method provided in an embodiment of the present invention, which specifically includes the following steps:
s301, acquiring initial data, and storing the initial data in a hard disk in a classified manner, wherein the initial data comprises data types and data amounts.
S302, acquiring the data quantity of initial data with the same data type in the hard disk.
When the initial data of the plurality of data types is stored in the storage area of the hard disk, the data volume of the initial data of the same data type can be obtained, so as to obtain the data volumes of the initial data of the various data types, for example: the data size of face data of the blacklisted person is 1 ten thousand.
S303, judging whether the data quantity of the initial data with the same data type reaches a preset data quantity threshold value.
The determination of the data amount may be a determination of the data amount of initial data of each data type and stored in the hard disk storage area. For example: the data types are 30 ten thousand and 20 ten thousand of initial data corresponding to the face of the blacklisted person and the vehicle information of the blacklisted vehicle in sequence respectively, and the preset number threshold is 25 ten thousand.
And S304, if the data quantity reaches the preset data quantity threshold, the initial data reaching the preset data quantity threshold are transferred to a video memory of the image processor for classified storage.
When the data quantity corresponding to the initial data of the same data type reaches a preset data quantity threshold, the initial data exceeding the preset data quantity threshold stored in the storage area of the hard disk can be transferred to the GPU video memory for storage. The above-mentioned transfer to GPU video memory to make classification storage can be that several storage areas are set in the video memory, and the initial data of different data types are respectively stored in different storage areas. When the data to be retrieved is obtained and needs to be retrieved in the GPU video memory, the initial data with the same data type as the data to be retrieved can be directly searched based on the different data types, and the comparison of the new similarity is completed, so that the initial data with the different data types from the data to be retrieved in the GPU video memory is not retrieved.
As a possible implementation mode, considering the characteristic that the GPU has expansibility, the GPU display card resource can be increased, and initial data can be uniformly distributed on each card. In this way, if there are k cards and n data, only k/n data need to be retrieved on each card, and after n cards perform similarity comparison, topN initial data with higher similarity are returned to the mobile terminal from which the user sends the data to be retrieved, for example: the method comprises the steps that 30 ten thousand initial data are arranged in total, the first video memory stores 10 ten thousand initial data, the second video memory stores 15 ten thousand initial data, the third video memory stores 5 ten thousand initial data, and if the data to be searched is the face of the blacklisted person, 10/30 data searching and similarity judging can be carried out in the first video memory. Of course, a plurality of video memories of the same data type may be provided, for example: the first video memory area and the second video memory area both store initial data of the same data type. Therefore, the speed of searching and similarity calculation can be increased, time consumption is reduced, and the data processing efficiency is improved.
S305, acquiring data to be retrieved, and judging whether initial data with the same data type as the data to be retrieved is stored in a video memory of the image processor.
S306, if the data to be searched and the initial data are compared in the video memory of the image processor, and the initial data with the similarity meeting the preset condition are used as target search data.
Optionally, the initial data includes an initial data feature value, and the step S304 may specifically include:
applying for a reserved memory area in a video memory of the image processor based on a data amount of the initial data reaching a preset data amount threshold.
After the data amount of the initial data reaching the preset data amount threshold is obtained, a reserved storage area larger than or equal to the data amount of the initial data reaching the preset data amount threshold may be applied to the GPU video memory, for example: the storage space occupied by the data volume transferred to the GPU is 2GB, and the reserved storage area of the application is 2.2GB. Therefore, enough space for storing the initial data with the data quantity reaching the preset data quantity threshold value can be ensured to be transferred to the reserved storage area of the GPU video memory.
And acquiring initial data characteristic values corresponding to initial data which are the same in data type and reach a preset data quantity threshold.
Each initial data may have a corresponding initial data characteristic value, where the initial data characteristic value can represent a characteristic of the initial data, and is fixed or variable-length semi-structured data formed by extracting specific information in image information. The initial data characteristic value of each data in the initial data with the same data type is obtained, the calculation speed can be increased by storing the initial data characteristic value, meanwhile, the space is saved, and the utilization rate of the GPU video memory area is improved.
And storing the initial data with the same data type and up to the preset data quantity threshold value and the initial data characteristic value in a reserved storage area of a video memory of the image processor.
The reserved storage area of the GPU video memory is set according to the space occupied by the initial data with the data quantity exceeding the preset data quantity threshold value, and the initial data characteristic value is stored in the reserved storage area corresponding to the GPU video memory, so that the storage area can be effectively and orderly used, and random storage is avoided.
Optionally, the step S305 may specifically include:
and receiving a request to be searched, wherein the request to be searched comprises data to be searched.
The user can send a request to be searched through the mobile terminal, and after the system receives the request to be searched, the system can respond to the request, namely, the process of searching and calculating in the initial data, and finally, a target search data is returned to the mobile terminal used by the user.
Extracting a characteristic value of the data to be searched, and identifying the data type of the data to be searched.
The data types of the data to be retrieved may also include the data types of the face of the blacklisted person, the human body of the blacklisted person, and the like described in the initial data. The data to be retrieved may be image data, and the image data may include information such as a person, a vehicle, and an action issued by the person or the vehicle, a scene in which the person or the vehicle is located (street, mall, office building, school, etc.). The data characteristic value of the data to be searched is extracted and can be used for comparing with the initial data characteristic value of the initial data, and the time for calculating the similarity can be shortened and the searching efficiency is improved through the comparison of the data characteristic value.
It is determined whether initial data having the same data type as the data to be retrieved is stored in a memory of the image processor.
The data type of the data to be searched can be searched before the similarity calculation, and after the storage area where the data type of the data to be searched is located is searched, the data type of the data to be searched is used as a target storage area for searching, for example: the target storage area is a hard disk, and initial data in the video memory of the GPU can not be involved by loading the initial data into the memory in the hard disk and performing similarity calculation; if the target storage area is the video memory of the GPU, similarity calculation is carried out in the video memory, and initial data in the hard disk cannot be involved. In this way, the initial data of different data types are not involved, and the retrieval time can be quickened.
If yes, searching a target storage area where initial data with the same data type as the data to be searched is located in a video memory of the image processor.
The target storage area may refer to a storage area in which initial data is the same as data type of data to be retrieved. If the initial data with the same data type as the data to be searched is found in the video memory of the GPU, a storage area of the initial data with the same data type in the video memory can be searched, after the initial data with the same data type is found, the storage area is the target storage area, and the search of the target storage area can facilitate the extraction of the initial data with the same data type and the similarity calculation of the initial data with the data to be searched.
Based on the characteristic value of the data to be searched, performing violent search in a target storage area to obtain an initial data characteristic value of initial data corresponding to the characteristic value of the data to be searched, performing similarity calculation on the characteristic value of the data to be searched and the initial data characteristic value, and taking the initial data with similarity meeting a preset condition as target search data.
The violent search may be to traverse the search in the hard disk/GPU video memory with the feature value and the data type of the data to be searched as the search conditions, and filter out the initial data which does not meet the search conditions, so as to obtain the target search data. The target search data may be one data or a set of data, for example: the data amount of the target retrieval data is 100.
If the target memory area is the GPU video memory, the similarity calculation may be performed by loading the initial data in the reserved memory space into the calculation space, and performing the similarity calculation in the calculation space of the GPU video memory. If the target storage area is a hard disk, the similarity calculation may be performed by loading initial data in the hard disk into a memory and performing calculation in the memory.
The similarity may be calculated by determining the distance between the feature value of the data to be retrieved and the space vector formed by the feature value of the initial data in a plurality of dimensions. The smaller the distance between two space vectors of the same dimension, the higher the similarity can be represented; the larger the distance, the lower the similarity. After the similarity calculation is completed, the GPU video memory or the hard disk memory can return N initial data with highest similarity as target retrieval data to the mobile terminal which sends out the request to be retrieved.
In the embodiment of the invention, after initial data are acquired and stored in the hard disk, whether the data amount corresponding to the initial data with the same data type reaches a preset data amount threshold value is judged, and the reached part is transferred from the hard disk to the GPU video memory for classified storage. When the data to be searched is obtained, if the storage area where the data type of the data to be searched is the GPU video memory, searching and similarity calculation are directly carried out in the GPU video memory, and the obtained similarity topN target search data are returned to the mobile terminal which sends the request to be searched. In this way, the initial data exceeding the preset data quantity threshold value is transferred to the GPU video memory, the data types with the same type are directly searched and calculated in the GPU video memory, the data processing capacity of the GPU is high, the concurrency characteristic is high, the searching speed is improved, the time consumption is reduced, and the utilization rate of the GPU video memory and the data processing efficiency are improved.
As shown in fig. 4, fig. 4 is a flowchart of another image data processing method provided in an embodiment of the present invention, which specifically includes the following steps:
s401, acquiring initial data, and storing the initial data in a hard disk in a classified manner, wherein the initial data comprises data types and data amounts.
S402, the initial data which are the same in data type and have the data quantity reaching a preset data quantity threshold value in the hard disk are transferred to a video memory of the image processor for classified storage.
S403, acquiring data to be searched, and judging whether initial data with the same data type as the data to be searched is stored in a video memory of the image processor.
S404, if the data to be searched and the initial data are compared in the video memory of the image processor, and the initial data with the similarity meeting the preset condition is used as target search data.
S405, deleting the initial data stored in the video memory of the image processor.
The data retrieval process needs the support of the initial data continuously, so that the initial data can be updated. When a new data type needs to be recorded, the updated data needs to be stored by a sufficient storage area, however, the storage area of the GPU video memory is limited, if the storage area in the GPU video memory is insufficient to store the initial data which is recorded subsequently, the initial data in the GPU video memory needs to be deleted, so that enough storage space is saved, for example: and deleting the initial data of the face of which the data type is the blacklisted person before 5 months. Of course, the initial data of what data type is specifically deleted can be freely selected, and the deleting condition can be set by itself, which is not limited in the embodiment of the present invention.
The deletion may be performed on all the initial data of the same data type, or may be performed on some of the initial data of a plurality of the same data types.
S406, judging whether the data quantity corresponding to the residual initial data with the same data type in the video memory of the image processor exceeds a preset data quantity threshold value.
After deleting the initial data in the GPU video memory, there are remaining initial data in the GPU video memory, where the remaining initial data may be undeleted data that may be used for retrieval and similarity calculation, for example: the data type is deleted. Judging whether the data quantity of the residual initial data of the same data type exceeds a preset data quantity threshold value, and facilitating storage conversion of the residual initial data.
S407, if the data type of the residual initial data in the video memory of the image processor is not exceeded, transferring the residual initial data with the same data type in the hard disk, and simultaneously clearing the transferred residual initial data in the video memory of the image processor.
When the data size of the remaining initial data of the same data type is still greater than or equal to the preset data size threshold, the remaining initial data of the same data type can be continuously stored in the GPU video memory.
As a possible embodiment, if the remaining initial data of the same type of data that is continuously stored in the GPU video memory occupies multiple memory areas of the video memory before being deleted, the remaining initial data in the multiple memory areas may be merged and stored after being partially deleted, for example: the storage area A is 3GB, 1GB of initial data is deleted, the remaining initial data is 2GB, the storage area B is 2GB, the remaining initial data is 1GB, the data types of the storage area A and the storage area B are consistent, and the remaining initial data in the storage area B can be transferred to the storage area A for storage. After merging, the initial data in one transferred storage area can be deleted, so that repeated existence is avoided, and the storage space is occupied.
When the data quantity of the residual initial data with the same data type is smaller than a preset data quantity threshold after being deleted, the residual initial data with the same data type in the GPU video memory is transferred to the memory of the hard disk, and when a request to be searched is received, the data type of the data to be searched is identified to be the same as the data type of the initial data transferred to the memory of the hard disk, and the searching and the similarity calculation can be directly performed in the hard disk. The above-mentioned clearing the remaining initial data transferred in the GPU video memory may be clearing after all the remaining initial data are transferred; or may be cleared directly after a remaining initial data transfer is completed. The method has the advantages that the transferred residual initial data are cleared, so that space is reserved for storing newly input data, the high availability of GPU video memory is guaranteed, and the condition of insufficient storage space is avoided.
In the embodiment of the invention, because the initial data is continuously recorded, the initial data in the GPU video memory is deleted to ensure the high availability of the GPU video memory, the initial data with the data quantity exceeding the preset data quantity threshold value is transferred to the hard disk for storage, and the retrieval and the similarity calculation are directly carried out in the hard disk when the data to be retrieved with the same data type is detected subsequently, so that the high availability of the GPU video memory is ensured.
As shown in fig. 5, fig. 5 is a schematic structural diagram of an apparatus for processing image data according to an embodiment of the present invention, where the apparatus specifically includes:
the acquiring module 501 is configured to acquire initial data, and store the initial data in a hard disk in a classified manner, where the initial data includes a data type and a data amount;
the transfer module 502 is configured to transfer initial data in the hard disk, where the data types are the same and the data amount reaches a preset data amount threshold, to a video memory of the image processor for classification storage;
a judging module 503, configured to obtain data to be retrieved, and judge whether initial data having the same data type as the data to be retrieved is stored in a video memory of the image processor;
and the calculating module 504 is configured to, if the data exists, perform similarity comparison on the data to be retrieved and the initial data in the video memory of the image processor, and take the initial data with the similarity meeting the preset condition as target retrieval data.
Optionally, the data type includes a data object and a data attribute, as shown in fig. 6, the obtaining module 501 includes:
a first recognition unit 5011 for recognizing a data type of the initial data;
the storage unit 5012 is configured to combine the initial data with the same data type, store the initial data with the same data type in the same hard disk, where the hard disk includes a plurality of hard disk storage areas, and each hard disk storage area stores the initial data with one data type.
Optionally, as shown in fig. 7, the dump module 502 includes:
an obtaining unit 5021, configured to obtain a data amount of initial data with the same data type in the hard disk;
a first determining unit 5022, configured to determine whether the data amount of the initial data with the same data type reaches a preset data amount threshold;
and the transferring unit 5023 is configured to transfer the initial data reaching the preset data amount threshold to the video memory of the image processor for classification storage if the initial data reaches the preset data amount threshold.
Optionally, the initial data includes an initial data feature value, as shown in fig. 8, and the dump unit 5023 includes:
an application subunit 50231, configured to apply for a reserved memory area in the video memory of the image processor based on the data amount of the initial data reaching the preset data amount threshold;
The obtaining subunit 50232 is configured to obtain an initial data characteristic value corresponding to initial data with the same data type and up to a preset data amount threshold;
the storage subunit 50233 is configured to store, in a reserved storage area of the video memory of the image processor, initial data that has the same data type and reaches a preset data amount threshold, and an initial data feature value.
Optionally, the storage subunit 50233 is further configured to mark initial data in the reserved storage area to obtain marked data, where the marked data includes a marked data type;
the storage subunit 50233 is further configured to store the tag data with the same tag data type in the reserved storage area in the video memory of the image processor.
Optionally, as shown in fig. 9, the determining module 503 includes:
a receiving unit 5031, configured to receive a request to be retrieved, where the request to be retrieved includes data to be retrieved;
a second identifying unit 5032, configured to extract a feature value of data to be retrieved, and identify a data type of the data to be retrieved;
a second judging unit 5033 for judging whether the initial data having the same data type as the data to be retrieved is stored in the video memory of the image processor;
The searching unit 5034 is configured to search, if so, a target storage area in which initial data having the same data type as the data to be retrieved is located in a video memory of the image processor;
the searching unit 5035 is configured to perform violent searching in the target storage area based on the feature value of the data to be searched to obtain an initial data feature value of initial data corresponding to the feature value of the data to be searched, perform similarity calculation on the feature value of the data to be searched and the feature value of the initial data, and use the initial data with similarity satisfying a preset condition as target searching data.
Optionally, as shown in fig. 10, the apparatus further includes:
a deleting module 505, configured to delete the initial data stored in the video memory of the image processor;
the judging module 503 is further configured to judge whether a data amount corresponding to the remaining initial data with the same data type in the video memory of the image processor exceeds a preset data amount threshold;
and the transferring module 506 is configured to transfer the remaining initial data with the same data type in the video memory of the image processor to the hard disk if the remaining initial data is not exceeded, and clear the transferred remaining initial data in the video memory of the image processor.
The image data processing device provided by the embodiment of the invention can realize each process and the same beneficial effects realized by the image data processing method in any method embodiment, and in order to avoid repetition, the description is omitted.
As shown in fig. 11, which is a schematic structural diagram of an electronic device according to an embodiment of the present invention, an electronic device 1000 includes: the memory 1102, the processor 1101, the network interface 1103, and a computer program stored on the memory 1102 and executable on the processor 1101 are communicatively connected to each other through a system bus. It should be noted that only electronic device 1000 having components 1101-1103 is shown in the figures, but it should be understood that not all of the illustrated components need be implemented and that more or fewer components may be implemented instead. It will be appreciated by those skilled in the art that the electronic device 1000 herein is a device capable of automatically performing numerical calculation and/or information processing according to instructions set or stored in advance, and its hardware includes, but is not limited to, a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a Programmable gate array (FPGA), a digital processor (Digital Signal Processor, DSP), an embedded device, and the like.
The electronic device 1000 may be a computing device such as a desktop computer, a notebook computer, a palm computer, a cloud server, etc. The electronic device 1000 may interact with a customer by way of a keyboard, mouse, remote control, touch pad, or voice control device.
The processor 1101 may be, among other things, a controller, microcontroller, microprocessor, or other data processing chip in some embodiments. The processor 1101 is typically used to control the overall operation of the computer device.
The processor 1101 is configured to call a computer program stored in the memory 1102, and perform the following steps:
acquiring initial data, and storing the initial data in a hard disk in a classified manner, wherein the initial data comprises a data type and a data amount;
the method comprises the steps of transferring initial data which are the same in data type and have data quantity reaching a preset data quantity threshold value in a hard disk into a video memory of an image processor for classified storage;
acquiring data to be searched, and judging whether initial data with the same data type as the data to be searched is stored in a video memory of an image processor or not;
if the data to be searched and the initial data are compared in the video memory of the image processor, and the initial data with the similarity meeting the preset condition is used as target search data.
Optionally, the step performed by the processor 1101 of storing the initial data classification in the hard disk includes:
identifying a data type of the initial data;
combining the initial data with the same data type, storing the initial data with the same data type in the same hard disk, wherein the hard disk comprises a plurality of hard disk storage areas, and each hard disk storage area stores the initial data with one data type.
Optionally, the step of transferring the initial data, which is executed by the processor 1101 and has the same data type and has a data amount reaching the preset data amount threshold, to the video memory of the image processor for classified storage includes:
acquiring the data quantity of initial data with the same data type in a hard disk;
judging whether the data quantity of initial data with the same data type reaches a preset data quantity threshold value or not;
and if the data quantity reaches the preset data quantity threshold, transferring the initial data which reaches the preset data quantity threshold to a video memory of the image processor for classified storage.
Optionally, the step of transferring the initial data, which reaches the preset data amount threshold, to the video memory of the image processor for classification storage includes:
applying for a reserved memory area in a video memory of the image processor based on a data amount of initial data reaching a preset data amount threshold;
acquiring initial data characteristic values corresponding to initial data which are the same in data type and reach a preset data quantity threshold;
and storing the initial data with the same data type and up to the preset data quantity threshold value and the initial data characteristic value in a reserved storage area of a video memory of the image processor.
Optionally, the step of storing the initial data having the same data type and reaching the preset data amount threshold and the initial data characteristic value in the reserved memory area of the video memory of the image processor performed by the processor 1101 includes:
marking the initial data in the reserved storage area to obtain marking data, wherein the marking data comprises marking data types;
and respectively storing the mark data with the same mark data type in a reserved storage area in a video memory of the image processor.
Optionally, the step of acquiring the data to be retrieved performed by the processor 1101, and determining whether the initial data having the same data type as the data to be retrieved is stored in the video memory of the image processor includes:
receiving a request to be searched, wherein the request to be searched comprises data to be searched;
extracting a characteristic value of the data to be searched, and identifying the data type of the data to be searched;
judging whether initial data with the same data type as the data to be retrieved is stored in a video memory of an image processor or not;
if yes, searching a target storage area where initial data with the same data type as the data to be searched is located in a video memory of the image processor;
based on the characteristic value of the data to be searched, performing violent search in a target storage area to obtain an initial data characteristic value of initial data corresponding to the characteristic value of the data to be searched, performing similarity calculation on the characteristic value of the data to be searched and the initial data characteristic value, and taking the initial data with similarity meeting a preset condition as target search data.
Optionally, the processor 1101 is further configured to perform the following steps:
deleting initial data stored in a memory of the image processor;
judging whether the data quantity of the residual initial data with the same data type in the video memory of the image processor exceeds a preset data quantity threshold value or not;
if the data type of the residual initial data in the video memory of the image processor is not exceeded, transferring the residual initial data with the same data type in the video memory of the image processor to the hard disk, and simultaneously clearing the transferred residual initial data in the video memory of the image processor.
The memory 1102 includes at least one type of readable storage medium including flash memory, hard disk, multimedia card, card memory (e.g., SD or DX memory, etc.), random Access Memory (RAM), static Random Access Memory (SRAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), programmable Read Only Memory (PROM), magnetic memory, magnetic disk, optical disk, etc.
The electronic device 1000 provided in the embodiment of the present invention can implement each implementation manner in the embodiment of the processing method of image data, and corresponding beneficial effects, and in order to avoid repetition, no description is repeated here.
The embodiment of the present invention further provides a computer readable storage medium, on which a computer program is stored, where the computer program when executed by the processor 1101 implements each process of the image data processing method provided by the embodiment of the present invention, and the same technical effects can be achieved, so that repetition is avoided, and no redundant description is provided herein.
Those skilled in the art will appreciate that the processes implementing all or part of the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, and the program may be stored in a computer readable storage medium, and the program may include the processes of the embodiments of the methods as above when executed. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited in order and may be performed in other orders, unless explicitly stated herein. Moreover, at least some of the steps in the flowcharts of the figures may include a plurality of sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, the order of their execution not necessarily being sequential, but may be performed in turn or alternately with other steps or at least a portion of the other steps or stages.
The terms "comprising" and "having" and any variations thereof in the description and claims of the present application and in the description of the figures above are intended to cover non-exclusive inclusions. The terms first, second and the like in the description and in the claims or in the above-described figures, are used for distinguishing between different objects and not necessarily for describing a sequential or chronological order. Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.
The foregoing disclosure is illustrative of the present invention and is not to be construed as limiting the scope of the invention, which is defined by the appended claims.
Claims (10)
1. A method of processing image data, comprising the steps of:
Acquiring initial data, and storing the initial data in a hard disk in a classified manner, wherein the initial data comprises a data type and a data volume;
the method comprises the steps of transferring initial data which are the same in data type and have data quantity reaching a preset data quantity threshold value in a hard disk into a video memory of an image processor for classified storage;
acquiring data to be retrieved, and judging whether the initial data with the same data type as the data to be retrieved is stored in a video memory of the image processor;
if so, comparing the similarity between the data to be searched and the initial data in the video memory of the image processor, and taking the initial data with the similarity meeting the preset condition as target search data.
2. The method of processing image data according to claim 1, wherein the step of storing the initial data classification in a hard disk includes:
identifying a data type of the initial data;
combining the initial data with the same data type, and storing the initial data with the same data type in the same hard disk, wherein the hard disk comprises a plurality of hard disk storage areas, and each hard disk storage area stores the initial data with one data type.
3. The method for processing image data according to claim 1, wherein the step of transferring the initial data of the same data type and having the data amount reaching the preset data amount threshold value to the video memory of the image processor for classification storage comprises:
acquiring the data quantity of initial data with the same data type in the hard disk;
judging whether the data quantity of the initial data with the same data type reaches a preset data quantity threshold value or not;
and if so, transferring the initial data reaching the preset data quantity threshold value to a video memory of the image processor for classified storage.
4. The method for processing image data according to claim 3, wherein the initial data includes an initial data feature value, and the step of transferring the initial data reaching the preset data amount threshold to a memory of the image processor for classification storage includes:
applying for a reserved storage area in a video memory of the image processor based on the data amount of the initial data reaching the preset data amount threshold;
acquiring initial data characteristic values corresponding to the initial data which are the same in data type and reach the preset data quantity threshold;
And storing the initial data with the same data type and up to the preset data quantity threshold value and the initial data characteristic value in the reserved storage area of the video memory of the image processor.
5. The method of processing image data according to claim 4, wherein the step of storing the initial data of the same data type and up to the preset data amount threshold and the initial data characteristic value in the reserved memory area of the video memory of the image processor includes:
marking the initial data in the reserved storage area to obtain marking data, wherein the marking data comprises marking data types;
and respectively storing the marking data with the same marking data type in the reserved storage area in the video memory of the image processor.
6. The method of processing image data according to claim 4, wherein the step of acquiring data to be retrieved, and determining whether the initial data having the same data type as the data to be retrieved is stored in a memory of the image processor comprises:
receiving a request to be searched, wherein the request to be searched comprises the data to be searched;
Extracting a characteristic value of the data to be searched, and identifying a data type of the data to be searched;
judging whether initial data with the same data type as the data to be retrieved is stored in a video memory of the image processor or not;
if yes, searching a target storage area where initial data with the same data type as the data to be searched is located in a video memory of the image processor;
and carrying out violent retrieval in the target storage area based on the data characteristic value to be retrieved so as to obtain an initial data characteristic value corresponding to the data characteristic value to be retrieved, carrying out similarity calculation on the data characteristic value to be retrieved and the initial data characteristic value, and taking the initial data with the similarity meeting a preset condition as target retrieval data.
7. The method of processing image data according to claim 1, wherein the method further comprises:
deleting initial data stored in a memory of the image processor;
judging whether the data quantity of the residual initial data with the same data type in the video memory of the image processor exceeds the preset data quantity threshold value or not;
and if the data type is not exceeded, transferring the residual initial data with the same data type in the video memory of the image processor to a hard disk, and simultaneously clearing the transferred residual initial data in the video memory of the image processor.
8. An image data processing apparatus, comprising:
the acquisition module is used for acquiring initial data, storing the initial data in a hard disk in a classified manner, wherein the initial data comprises a data type and a data amount;
the transfer module is used for transferring the initial data which are the same in data type and have the data quantity reaching the preset data quantity threshold value in the hard disk into the video memory of the image processor for classified storage;
the judging module is used for acquiring data to be searched and judging whether the initial data with the same data type as the data to be searched is stored in a video memory of the image processor or not;
and the calculation module is used for comparing the similarity between the data to be retrieved and the initial data in the video memory of the image processor if the data to be retrieved exist, and taking the initial data with the similarity meeting the preset condition as target retrieval data.
9. An electronic device, comprising: memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the method of processing image data according to any one of claims 1 to 7 when the computer program is executed.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps in the method of processing image data according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911338762.1A CN113094529B (en) | 2019-12-23 | 2019-12-23 | Image data processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911338762.1A CN113094529B (en) | 2019-12-23 | 2019-12-23 | Image data processing method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113094529A CN113094529A (en) | 2021-07-09 |
| CN113094529B true CN113094529B (en) | 2024-01-05 |
Family
ID=76662918
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911338762.1A Active CN113094529B (en) | 2019-12-23 | 2019-12-23 | Image data processing method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113094529B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114722229B (en) * | 2022-03-22 | 2025-09-23 | 京东方科技集团股份有限公司 | Face recognition method and face recognition device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018077292A1 (en) * | 2016-10-28 | 2018-05-03 | 北京市商汤科技开发有限公司 | Data processing method and system, electronic device |
| CN110321448A (en) * | 2019-06-27 | 2019-10-11 | 腾讯科技（深圳）有限公司 | A kind of image search method, device and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8886206B2 (en) * | 2009-05-01 | 2014-11-11 | Digimarc Corporation | Methods and systems for content processing |
-
2019
- 2019-12-23 CN CN201911338762.1A patent/CN113094529B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018077292A1 (en) * | 2016-10-28 | 2018-05-03 | 北京市商汤科技开发有限公司 | Data processing method and system, electronic device |
| CN110321448A (en) * | 2019-06-27 | 2019-10-11 | 腾讯科技（深圳）有限公司 | A kind of image search method, device and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 基于改进的分布式K-Means特征聚类的海量场景图像检索;崔红艳;曹建芳;;计算机应用与软件(06);全文 * |
| 纤支镜检查实时图文信息报告系统;陈石平, 刘立亚;医疗设备信息(05);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113094529A (en) | 2021-07-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10885100B2 (en) | Thumbnail-based image sharing method and terminal | |
| CN108121816A (en) | Picture classification method, device, storage medium and electronic equipment | |
| CN110321448B (en) | Image retrieval method, device and storage medium | |
| CN108288208B (en) | Display object determination method, device, medium and equipment based on image content | |
| CN105224528B (en) | Big data processing method and device based on graph computing | |
| WO2021027595A1 (en) | User portrait generation method and apparatus, computer device, and computer-readable storage medium | |
| CN103399896A (en) | Method and system for recognizing association relationships among users | |
| WO2021072592A1 (en) | Connected domain analysis method, data processing device, and computer readable storage medium | |
| CN112052251B (en) | Target data updating method and related device, equipment and storage medium | |
| CN111177450B (en) | Image retrieval cloud identification method and system and computer readable storage medium | |
| CN112597363B (en) | User processing method and device, electronic equipment and storage medium | |
| CN113094529B (en) | Image data processing method and device, electronic equipment and storage medium | |
| CN113887367A (en) | An identity filing method, device, electronic device and storage medium | |
| CN111159443B (en) | Image characteristic value searching method and device and electronic equipment | |
| CN110046632B (en) | Model training method and device | |
| CN113360688B (en) | Method, device and system for constructing information base | |
| CN117369731B (en) | Data reduction processing method, device, equipment and medium | |
| CN113407107B (en) | Data storage method, device and equipment | |
| CN114332714B (en) | Data management method and device and computer readable storage medium | |
| EP4136580A1 (en) | Target re-identification method, network training method thereof, and related device | |
| CN115934980B (en) | Layout graph search processing method and device, equipment and storage medium | |
| CN114972229B (en) | Material type-based damage detection method and device, electronic equipment and medium | |
| CN114360011B (en) | Image recognition method, device, equipment and medium | |
| JP7180769B2 (en) | Data management device, control method, and storage medium | |
| KR20190013057A (en) | Method and apparatus for automatically classifying mobile application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |