CN117315525A - Image recognition effect inspection method, device and storage medium - Google Patents

Abstract

The invention relates to an image recognition effect checking method, an image recognition effect checking device and a storage medium, wherein the method comprises the following steps: setting a display mode of the display module; setting a display mode of the transformation module; setting an interference mode of the interference module; the image acquisition device is arranged, so that the image acquisition device can acquire test images or test videos displayed by each display device and each conversion device; the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition; the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

Description

Image recognition effect inspection method, device and storage medium

Technical Field

The invention belongs to the field of image recognition testing, and particularly relates to an image recognition effect checking method, an image recognition effect checking device and a storage medium.

Background

At present, when an intelligent recognition algorithm and intelligent recognition equipment are inspected, a large number of sample tests are often required to be carried out so as to evaluate the recognition effect. The method is closest to the actual effect, a large amount of manpower and material resources are input to verify the actual condition, the operation effect is clear at a glance, but the method is not cost-effective, not only is the waste of resources caused, but also the double increase of the cost is caused once the equipment is adjusted and changed.

The current common detection methods are: firstly, an identification algorithm is developed, a plurality of photos/videos are collected to be used as a test set for verifying the identification effect, then the test set is read through a program and sent to an algorithm module, the identified result is output through algorithm processing, and then tuning or integration is carried out with other modules. Testing of the development phase has been done in a matter of paragraph. As a part of the defects, a tester can find that, firstly, for the algorithm module, certain photos are taken all the time, the binary data read by a computer is unchanged, as long as the test set is unchanged, the data source acquired by the recognition algorithm is unchanged, and if the data set is not comprehensive enough, the actual use effect is affected due to weak representativeness. Secondly, the developer often tests the algorithm separately and then integrates the algorithm with other modules, and when all the modules are deployed in the hardware device, whether the problem is generated in the actual recognition process is unknown at this stage, which is of course important concern to the tester.

Disclosure of Invention

The invention provides an image recognition effect checking method, an image recognition effect checking device and a storage medium, and aims to at least solve one of the technical problems in the prior art.

The technical scheme of the invention relates to an image recognition effect checking device, which is characterized by comprising:

the control module is used for controlling the image recognition effect checking device;

the image test set module is used for collecting and storing various types of test images and is connected with the control module;

the display module is used for displaying images or videos of the image test set module and is connected with the control module;

the transformation module is used for displaying images or videos of the image test set which are subjected to transformation processing by the control module, the number of the transformation modules is at least one, and the transformation module is connected with the control module;

the interference module is used for generating environmental interference and is adjacent to or opposite to the transformation module;

the intelligent recognition module is used for recognizing information of images or videos, the intelligent recognition module comprises an image acquisition device and a recognition device which are sequentially connected, the image acquisition device is used for acquiring the display module, the conversion module and the images or videos interfered by the interference module, the recognition device is used for recognizing the images or videos acquired by the image acquisition device, the image acquisition device is connected with the recognition device through a cable or a network, and the intelligent recognition module is connected with the control module;

and the result summarizing module is used for summarizing and analyzing the judging result of the intelligent identification module and outputting a result summarizing file, and the intelligent identification module and the control module are respectively connected with the result summarizing module.

Further, the method also comprises an application program, wherein the application program executes the following steps:

s100, setting a display mode of the display module, wherein the display module invokes a test image or video from the image test set module through the control module for display;

s200, setting a display mode of the transformation module, wherein the transformation module is used for displaying images or videos of the image test set transformed by the control module, and the control module invokes test images or test videos from the image test set module;

s300, setting an interference mode of the interference module, wherein the interference module is used for interfering the environments of the display module and the transformation module;

s400, setting the image acquisition device to enable the image acquisition device to acquire test images or test videos displayed by each display device and each conversion device;

s500, the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition;

s600, the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

Further, step S100 includes:

s110, setting and calling a range of a test image or a test video in the image test module, wherein the image test module comprises a plurality of types of test images and test videos;

s120, setting a display mode of the display module, wherein the display mode of the display module comprises the display time of each test set image, whether an irrelevant image is inserted or not and the display time of the irrelevant image;

s130, setting the number of times of cyclic display of the display module.

Further, step S200 includes:

s210, setting the number of the transformation modules;

s220, setting a conversion processing mode of each conversion module;

s230, according to the transformation processing mode of each transformation module, the control module generates a corresponding transformed test image or transformed test video based on the test image or the test video;

s240, setting a display mode of the transformation module, wherein the display mode of the transformation module comprises the display time of each transformed test image or transformed test video, whether an irrelevant image is inserted or not and the display time of the irrelevant image;

s250, setting the times of each transformation module for circularly displaying the transformed test image or the transformed test video.

Further, the transformation processing mode comprises a transformation image obtained after the image of the image test set is subjected to amplification processing, shrinkage processing, mirror image processing, inclination processing, stretching processing, concave-convex processing and illumination uneven processing.

Further, the interference module includes light source interference, specular interference, and obscuration interference.

Further, step S400 includes:

s410, setting each image acquisition device to be respectively aligned with the display device or the conversion module;

s420, if the network camera equipment exists, checking the connection state of the network camera equipment and the intelligent recognition module.

Further, step S600 includes:

s610, the intelligent recognition module recognizes the test image or the test video acquired by the image acquisition device based on the item of the data tag of the test image or the test video to obtain a recognition result;

s620, the intelligent recognition module compares whether the test result is correct one by one based on the data labels of the test image or the test video, and calculates the accuracy;

s630, storing and outputting the test result.

Further, the present invention also provides an image recognition effect inspection device, which includes:

the control module comprises a server;

the image test set module comprises a server and a network storage server, and is connected with the control module;

the display module comprises a display, and is connected with the control module;

the conversion module comprises a display, and is connected with the control module;

the interference module comprises a light source interference lamp, a mirror interference object and a shielding object;

the intelligent identification module comprises a server, and is connected with the control module;

and the result summarizing module comprises a server, and is connected with the control module.

The invention further provides an image recognition effect checking method, the image recognition effect checking device comprises the following steps:

s100, setting a display mode of the display module, wherein the display module invokes a test image or video from the image test set module through the control module for display;

s200, setting a display mode of the transformation module, wherein the transformation module is used for displaying images or videos of the image test set transformed by the control module, and the control module invokes test images or test videos from the image test set module;

s300, setting an interference mode of the interference module, wherein the interference module is used for interfering the environments of the display module and the transformation module;

s400, setting the image acquisition device to enable the image acquisition device to acquire test images or test videos displayed by each display device and each conversion device;

s500, the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition;

s600, the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

Further, the present invention also proposes a computer readable storage medium having stored thereon program instructions which, when executed by a processor, implement the method.

Compared with the prior art, the invention has the following characteristics:

the method can simulate the acquisition effect of the image in various real environments, simulate the interference factors in reality, automatically acquire, identify and compare the image or the converted image with the preset data label, thereby realizing the technical effect of automatically analyzing the accuracy of the identification program.

Drawings

FIG. 1 is a flow chart of an image recognition effect inspection method;

FIG. 2 is a flow chart of a display mode of the display module in the image recognition effect checking method;

FIG. 3 is a flowchart of a display mode of setting the transformation module in the image recognition effect checking method;

FIG. 4 is a flowchart of the image acquisition device in the image recognition effect inspection method;

FIG. 5 is a flow chart of an intelligent recognition module in an image recognition effect inspection method;

FIG. 6 is a schematic diagram of an image recognition effect inspection apparatus;

fig. 7 is a schematic diagram of an output test result of the intelligent recognition module in the image recognition effect checking method.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. 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.

The conception, specific structure, and technical effects produced by the present invention will be clearly and completely described below with reference to the embodiments and the drawings to fully understand the objects, aspects, and effects of the present invention.

It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly or indirectly fixed or connected to the other feature. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description presented herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any combination of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element of the same type from another. For example, a first element could also be termed a second element, and, similarly, a second element could also be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples, or exemplary language (e.g., "such as") provided herein, is intended merely to better illuminate embodiments of the invention and does not pose a limitation on the scope of the invention unless otherwise claimed. Further, as used herein, the industry term "pose" refers to the position and pose of an element relative to a spatial coordinate system.

Referring to fig. 1 to 7, an embodiment of the present invention provides an image recognition effect inspection apparatus, which is characterized in that, referring to fig. 6, the apparatus includes:

the control module is used for controlling the image recognition effect checking device;

the image test set module is used for collecting and storing various types of test images and is connected with the control module;

the display module is used for displaying images or videos of the image test set module and is connected with the control module;

the transformation module is used for displaying images or videos of the image test set which are subjected to transformation processing by the control module, the number of the transformation modules is at least one, and the transformation module is connected with the control module;

the interference module is used for generating environmental interference and is adjacent to or opposite to the transformation module;

the intelligent recognition module is used for recognizing information of images or videos, the intelligent recognition module comprises an image acquisition device and a recognition device which are sequentially connected, the image acquisition device is used for acquiring the display module, the conversion module and the images or videos interfered by the interference module, the recognition device is used for recognizing the images or videos acquired by the image acquisition device, the image acquisition device is connected with the recognition device through a cable or a network, and the intelligent recognition module is connected with the control module;

and the result summarizing module is used for summarizing and analyzing the judging result of the intelligent identification module and outputting a result summarizing file, and the intelligent identification module and the control module are respectively connected with the result summarizing module.

In some embodiments, the display module and the transformation module are displays such as a normal display, a hairtail screen, a curved screen, and the like.

Further, referring to fig. 1, an application program is further included, and the application program performs the following steps:

s100, setting a display mode of the display module, wherein the display module invokes a test image or video from the image test set module through the control module for display;

s200, setting a display mode of the transformation module, wherein the transformation module is used for displaying images or videos of the image test set transformed by the control module, and the control module invokes test images or test videos from the image test set module;

s300, setting an interference mode of the interference module, wherein the interference module is used for interfering the environments of the display module and the transformation module;

s400, setting the image acquisition device to enable the image acquisition device to acquire test images or test videos displayed by each display device and each conversion device;

s500, the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition;

s600, the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

Further, referring to fig. 1 and 2, step S100 includes:

s110, setting and calling a range of a test image or a test video in the image test module, wherein the image test module comprises a plurality of types of test images and test videos;

s120, setting a display mode of the display module, wherein the display mode of the display module comprises the display time of each test set image, whether an irrelevant image is inserted or not and the display time of the irrelevant image;

s130, setting the number of times of cyclic display of the display module.

In some embodiments, the test image or test video includes a face image video, a vehicle image video, a license plate image video, and a helmet image video; the data labels of the face image video comprise gender, age and expression; the data labels of the vehicle image video comprise vehicle types, vehicle colors and cab portraits; the data labels of the license plate image video comprise license plate numbers, license plate colors and license plate types.

In some embodiments, the display mode of the image on the display is set, and the display mode can be controlled by software programming on the control module, so that the image display area and the pause interval can be controlled in detail and strictly. When the setting interval is large, the image recognition effect can be tested, a large number of similar images (dissimilar images can be present) exist in the data set, and when the setting interval is small, the input video stream can be simulated, namely, the algorithm and the equipment supporting the video analysis can be tested, and the setting is easy to realize. If each image is set to stay for 40ms, when the images are circularly displayed, the frame rate which is commonly used in video can be 25 frames per second, and the value can be flexibly set, so that the method is helpful for the image recognition speed of a test algorithm in practice. In addition, according to different surrounding environments, the position, the angle and the illumination of the display module can be properly adjusted to obtain different display effects. When the display module is arranged, different backgrounds and shields can be configured at other places on the screen outside the image display area so as to construct different environmental interferences.

Further, referring to fig. 1 and 3, step S200 includes:

s210, setting the number of the transformation modules;

s220, setting a conversion processing mode of each conversion module;

s230, according to the transformation processing mode of each transformation module, the control module generates a corresponding transformed test image or transformed test video based on the test image or the test video;

s240, setting a display mode of the transformation module, wherein the display mode of the transformation module comprises the display time of each transformed test image or transformed test video, whether an irrelevant image is inserted or not and the display time of the irrelevant image;

s250, setting the times of each transformation module for circularly displaying the transformed test image or the transformed test video.

Specifically, the transformation module performs certain transformation on the image in the display module, and the effect after the transformation is displayed in the transformation module. In some embodiments, the image or video displayed by the transformation module may be an image or video of an original image or video after being transformed by the optical device, such as a plane mirror transformation, a convex mirror transformation, and the like. The number of the conversion modules is not limited to two, but two are preferable in consideration of the actual spatial position, imaging influence with each other, and the like.

After the transformation device is added, a larger visual angle range and different image special effects can be obtained by changing the position of the intelligent recognition device. If the device A is placed in a place opposite to the display to obtain a front view angle, the device B is placed in the display in parallel, in a small-angle inclination, in a large-angle inclination and partially beyond the display through a mechanical shaft, a supporting platform and the like to achieve different view angle ranges. Aiming at an intelligent identification device, the image identification data source is enlarged, meanwhile, if the number of the devices is increased, and the positions of the devices are reasonably placed, the identification effects of the same image under different angles can be compared, namely, the test efficiency is improved under a time unit. In addition, according to different surrounding environments, the position, the angle and the illumination of the conversion module can be properly adjusted to obtain different display effects. It is useful to perform a limit test of the wide angle support range.

Further, referring to fig. 1, the transformation processing mode includes a transformation image obtained by performing an enlargement process, a reduction process, a mirroring process, a tilting process, a stretching process, a concave-convex process, and an uneven illumination process on the image of the image test set.

Further, referring to fig. 1, the interference module includes light source interference, specular interference, and obscuration interference.

Further, referring to fig. 1 and 4, step S400 includes:

s410, setting each image acquisition device to be respectively aligned with the display device or the conversion module;

s420, if the network camera equipment exists, checking the connection state of the network camera equipment and the intelligent recognition module.

Specifically, for step S500, the device starts to identify, that is, before starting to run the testing process, a certain configuration may be performed on the device, including, but not limited to, the number of devices participating in the test, adjusting the location of the device, configuring the identification parameters of the device, and configuring the parameters of the device may be performed in the control module. In the testing process, the equipment can be changed at any time according to the change of the environment or the adjustment of the testing strategy, and the test can be selected to be suspended or not suspended at the moment.

Further, referring to fig. 1, the test image or test video includes a face image video, a vehicle image video, a license plate image video, and a helmet image video;

the data labels of the face image video comprise gender, age and expression;

the data labels of the vehicle image video comprise vehicle types, vehicle colors and cab portraits;

the data labels of the license plate image video comprise license plate numbers, license plate colors and license plate types.

Further, referring to fig. 1 and 5, step S600 includes:

s610, the intelligent recognition module recognizes the test image or the test video acquired by the image acquisition device based on the item of the data tag of the test image or the test video to obtain a recognition result;

s620, the intelligent recognition module compares whether the test result is correct one by one based on the data labels of the test image or the test video, and calculates the accuracy;

s630, storing and outputting the test result.

Specifically, the intelligent recognition module needs to send the recognition result to the result summarization module, and the sent information is determined by an algorithm embedded in the device, and according to the test target, detailed or abbreviated results can be sent, but the items marked in the data set are used for statistics. The results sent by the identification device are also formatted.

The result summarizing module is also controlled by the control module, the control module can temporarily store the identification result after receiving the result sent by the identification equipment to form a file format, automatically read the identification result file and compare with the standard result file after the test of the round is finished, automatically calculate and generate the correct rate and the false identification rate of the test, and can generate a result graph of a few times by comparing the results of the tests, thereby being convenient for observing the improvement effect and the running trend. The method can be realized easily by programming, and can be carried out when a certain number and a certain time are reached without summarizing the results after the test is completed. Meanwhile, the reserved original result file is convenient for development and test personnel to carry out manual secondary verification.

Through the working flow, the method can realize that after relevant modules are configured, a data set is prepared, and a test is started, the system can be completely handed over to automatically execute, and finally, a readable test result is generated.

Further, referring to fig. 6, the present invention also proposes an image recognition effect checking apparatus, including:

the control module comprises a server;

the image test set module comprises a server and a network storage server, and is connected with the control module;

the display module comprises a display, and is connected with the control module;

the conversion module comprises a display, and is connected with the control module;

the interference module comprises a light source interference lamp, a mirror interference object and a shielding object;

the intelligent identification module comprises a server, and is connected with the control module;

and the result summarizing module comprises a server, and is connected with the control module.

Further, referring to fig. 1, the present invention further provides an image recognition effect checking method, and the image recognition effect checking device, and the method includes the following steps:

s100, setting a display mode of the display module, wherein the display module invokes a test image or video from the image test set module through the control module for display;

s200, setting a display mode of the transformation module, wherein the transformation module is used for displaying images or videos of the image test set transformed by the control module, and the control module invokes test images or test videos from the image test set module;

s300, setting an interference mode of the interference module, wherein the interference module is used for interfering the environments of the display module and the transformation module;

s400, setting the image acquisition device to enable the image acquisition device to acquire test images or test videos displayed by each display device and each conversion device;

s500, the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition;

s600, the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

Further, referring to fig. 1, the present invention also proposes a computer readable storage medium having stored thereon program instructions which, when executed by a processor, implement the method.

Compared with the prior art, the invention has the following characteristics:

the method can simulate the acquisition effect of the image in various real environments, simulate the interference factors in reality, automatically acquire, identify and compare the image or the converted image with the preset data label, thereby realizing the technical effect of automatically analyzing the accuracy of the identification program.

In particular, the method comprises the steps of,

1) The test efficiency is improved, the test device is deployed once, the test device can be used for multiple times in the follow-up regression test, even if the test device is a non-developer, common test personnel can easily use the test device, the automation degree is high, and a large number of samples and test works with different visual angles can be processed at one time. By configuring the simulated real environment of the scene, the influence of the input of the test set to the identification equipment on the actual effect is avoided.

2) The test environment is stable, if the system is deployed in a place where the surrounding environment is not easy to change, the unified external factors of each test can be ensured, and a plurality of external interferences are reduced for the experiments of controlling the variable type.

3) The system structure has strong flexibility and expansibility, and each component which is independently distributed is shown in the system structure, if the system structure can be manufactured into a unified machine, and other components such as a light source, a light shield, infrared rays and the like are added, the system structure can be flexibly deployed and provide more testing scenes, and if the display equipment is improved, if flexible display is used, a more real use scene can be simulated. The selection and display modes of the data source have no strict requirements, and the flexible change is carried out according to software control, so that the method not only can be used for image recognition, but also can be applied to video analysis to a certain extent.

4) The system design of combining soft and hard is convenient in later maintenance, and under the condition that the test scheme is unchanged, the change of the test strategy only needs to adjust the relevant software configuration, so that the method is very friendly to novice testers.

In a specific embodiment, an algorithm for face recognition is tested for its recognition effect under low illumination, high light, and side Fang An viewing angles.

Material to be prepared first:

1) 100 front face photos are used as test sets, numbering is completed, and various pieces of information of each photo are stored in an excel table, wherein the excel table comprises different age groups, skin colors, partial glasses and the like;

2) Three intelligent recognition devices, namely recognition modules, can be provided with cameras or can be connected with the cameras through a network and the like to collect pictures;

3) Light supplementing lamps, plane mirrors, other mechanical components and the like required by the conversion module.

The method comprises the following steps:

1) Initializing an intelligent identification module: burning an algorithm to be tested into the identification equipment, adding the algorithm of face recognition into a firmware package, setting the output of the algorithm to be only a person name, and then using a tool to burn the firmware into the equipment; the front photo personnel information of the test set is then registered with each device so that the identification module has a "standard answer".

2) Setting a display module and a conversion module: the display module here uses a common flat panel display, displaying images of a test set every 3s, with a blank image interposed, the blank image being displayed for 1s. The front image is not used as a testing concern point. Then, plane mirrors are placed on two sides of the display module, wherein one plane mirror is provided with strong light, and the other plane mirror is partially shielded by a light shielding plate, so that in order to prevent strong light interference, the display module and the conversion module can be additionally staggered to reduce interference. And then a plane mirror is placed at a proper position such as the upper right corner, so that the plane mirror is obviously inclined relative to the picture in the display module. In this way, the display period of the transformation module is consistent with that of the display module, and the actual picture is close to the expected strong light, weak light and wide viewing angle, and then can be further corrected.

In a specific embodiment, this step is implemented by the following code:

3) Camera of intelligent identification module puts: the default recognition modules collect images from the cameras and then conduct intelligent recognition analysis.

Mode one: using a camera of the identification module;

mode two: the IPC image is obtained through a network, the IPC image is directly adopted, the three cameras are respectively opposite to the three transformed images and fixed, and then the image acquisition of three states is completed.

4) Summarizing and comparing results: in view of the fact that only face recognition is carried out at this time, the recognition module can record recognition results of each time, the recognition results of three devices are respectively uploaded and stored in the result summarizing module in a text document mode by the control module. After the round is finished, the result summarizing module compares the standard answers of excel and the recognition results of three recognition devices, namely the text documents, one by one, and then automatically calculates the accuracy rate under each scene.

In summary, the result summarizing module may be regarded as a folder storing "standard answers" and actual recognition results, and the control module may be regarded as an automated program for operating these files, and outputs the accuracy after each round of testing.

In a specific embodiment, this step is implemented by the following code:

referring to fig. 7, the output result includes 100 sets of data of numbers 1 to 100, wherein each set of data includes 3 test results, images or videos of 1 display module and 2 transform modules are respectively collected, and the comparison results 1 to 3 are respectively comparison results of the test results and the data tag, and are true if the test results are consistent with the data tag, and are false if the test results are inconsistent with the data tag. And finally, evaluating the accuracy of a test program by respectively calculating the accuracy of the comparison results 1 to 3.

It should be appreciated that the method steps in embodiments of the present invention may be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in non-transitory computer-readable memory. The method may use standard programming techniques. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.

Furthermore, the operations of the processes described herein may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes (or variations and/or combinations thereof) described herein may be performed under control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications), by hardware, or combinations thereof, collectively executing on one or more processors. The computer program includes a plurality of instructions executable by one or more processors.

Further, the method may be implemented in any type of computing platform operatively connected to a suitable computing platform, including, but not limited to, a personal computer, mini-computer, mainframe, workstation, network or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and so forth. Aspects of the invention may be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optical read and/or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer, which when read by a computer, is operable to configure and operate the computer to perform the processes described herein. Further, the machine readable code, or portions thereof, may be transmitted over a wired or wireless network. When such media includes instructions or programs that, in conjunction with a microprocessor or other data processor, implement the steps described above, the invention described herein includes these and other different types of non-transitory computer-readable storage media. The invention may also include the computer itself when programmed according to the methods and techniques of the present invention.

The computer program can be applied to the input data to perform the functions described herein, thereby converting the input data to generate output data that is stored to the non-volatile memory. The output information may also be applied to one or more output devices such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on a display.

The present invention is not limited to the above embodiments, but can be modified, equivalent, improved, etc. by the same means to achieve the technical effects of the present invention, which are included in the spirit and principle of the present invention. Various modifications and variations are possible in the technical solution and/or in the embodiments within the scope of the invention.

Claims (10)

1. An image recognition effect inspection apparatus, characterized in that the apparatus comprises:

the control module is used for controlling the image recognition effect checking device;

the image test set module is used for collecting and storing various types of test images and is connected with the control module;

the display module is used for displaying images or videos of the image test set module and is connected with the control module;

the transformation module is used for displaying images or videos of the image test set which are subjected to transformation processing by the control module, the number of the transformation modules is at least one, and the transformation module is connected with the control module;

the interference module is used for generating environmental interference and is adjacent to or opposite to the transformation module;

the intelligent recognition module is used for recognizing information of images or videos, the intelligent recognition module comprises an image acquisition device and a recognition device which are sequentially connected, the image acquisition device is used for acquiring the display module, the conversion module and the images or videos interfered by the interference module, the recognition device is used for recognizing the images or videos acquired by the image acquisition device, the image acquisition device is connected with the recognition device through a cable or a network, and the intelligent recognition module is connected with the control module;

and the result summarizing module is used for summarizing and analyzing the judging result of the intelligent identification module and outputting a result summarizing file, and the intelligent identification module and the control module are respectively connected with the result summarizing module.

2. The image recognition effect inspection apparatus according to claim 1, further comprising an application program that performs the steps of:

s100, setting a display mode of the display module, wherein the display module invokes a test image or video from the image test set module through the control module for display;

s200, setting a display mode of the transformation module, wherein the transformation module is used for displaying images or videos of the image test set transformed by the control module, and the control module invokes test images or test videos from the image test set module;

s300, setting an interference mode of the interference module, wherein the interference module is used for interfering the environments of the display module and the transformation module;

s400, setting the image acquisition device to enable the image acquisition device to acquire test images or test videos displayed by each display device and each conversion device;

s500, the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition;

s600, the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

3. The image recognition effect checking apparatus according to claim 2, wherein step S100 includes:

s110, setting and calling a range of a test image or a test video in the image test module, wherein the image test module comprises a plurality of types of test images and test videos;

s120, setting a display mode of the display module, wherein the display mode of the display module comprises the display time of each test set image, whether an irrelevant image is inserted or not and the display time of the irrelevant image;

s130, setting the number of times of cyclic display of the display module.

4. The image recognition effect checking apparatus according to claim 2, wherein step S200 includes:

s210, setting the number of the transformation modules;

s220, setting a conversion processing mode of each conversion module;

s230, according to the transformation processing mode of each transformation module, the control module generates a corresponding transformed test image or transformed test video based on the test image or the test video;

s240, setting a display mode of the transformation module, wherein the display mode of the transformation module comprises the display time of each transformed test image or transformed test video, whether an irrelevant image is inserted or not and the display time of the irrelevant image;

s250, setting the times of each transformation module for circularly displaying the transformed test image or the transformed test video.

5. The image recognition effect inspection apparatus according to claim 4, wherein the transformation processing means includes a transformed image obtained by performing an enlargement process, a reduction process, a mirroring process, a tilting process, a stretching process, a concavity and convexity process, and an uneven illumination process on the image of the image test set.

6. The image recognition effect checking apparatus according to claim 2, wherein step S400 includes:

s410, setting each image acquisition device to be respectively aligned with the display device or the conversion module;

s420, if the network camera equipment exists, checking the connection state of the network camera equipment and the intelligent recognition module.

7. The image recognition effect checking apparatus according to claim 2, wherein step S600 includes:

s610, the intelligent recognition module recognizes the test image or the test video acquired by the image acquisition device based on the item of the data tag of the test image or the test video to obtain a recognition result;

s620, the intelligent recognition module compares whether the test result is correct one by one based on the data labels of the test image or the test video, and calculates the accuracy;

s630, storing and outputting the test result.

8. The image recognition effect checking apparatus according to claim 1, wherein,

the control module comprises a server;

the image test set module comprises a server and a network storage server, and is connected with the control module;

the display module comprises a display, and is connected with the control module;

the conversion module comprises a display, and is connected with the control module;

the interference module comprises a light source interference lamp, a mirror interference object and a shielding object;

the intelligent identification module comprises a server, and is connected with the control module;

and the result summarizing module comprises a server, and is connected with the control module.

9. An image recognition effect checking method for the image recognition effect checking apparatus according to claim 1, the method comprising the steps of:

s100, setting a display mode of the display module, wherein the display module invokes a test image or video from the image test set module through the control module for display;

s200, setting a display mode of the transformation module, wherein the transformation module is used for displaying images or videos of the image test set transformed by the control module, and the control module invokes test images or test videos from the image test set module;

s300, setting an interference mode of the interference module, wherein the interference module is used for interfering the environments of the display module and the transformation module;

s400, setting the image acquisition device to enable the image acquisition device to acquire test images or test videos displayed by each display device and each conversion device;

s500, the control module controls the display module and the conversion module to start playing test images and test videos, and the intelligent recognition module controls the image acquisition devices to start image acquisition;

s600, the intelligent recognition module analyzes and gathers the images or videos acquired by the image acquisition device and outputs analysis results, wherein the analysis results comprise accuracy after comparison based on the data labels of the test images or the test videos.

10. A computer readable storage medium having stored thereon program instructions which, when executed by a processor, implement the method of any of claims 9.