CN114358166B - Multi-target positioning method based on self-adaptive k-means clustering - Google Patents
Multi-target positioning method based on self-adaptive k-means clustering Download PDFInfo
- Publication number
- CN114358166B CN114358166B CN202111634014.5A CN202111634014A CN114358166B CN 114358166 B CN114358166 B CN 114358166B CN 202111634014 A CN202111634014 A CN 202111634014A CN 114358166 B CN114358166 B CN 114358166B
- Authority
- CN
- China
- Prior art keywords
- matching
- feature
- feature points
- point
- coordinates
- 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
- 238000000034 method Methods 0.000 title claims abstract description 15
- 238000003064 k means clustering Methods 0.000 title claims abstract description 11
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 35
- 230000009466 transformation Effects 0.000 claims abstract description 10
- 238000005286 illumination Methods 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 241000282320 Panthera leo Species 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 239000002356 single layer Substances 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a multi-target positioning method based on self-adaptive k-means clustering, which combines a density peak clustering algorithm (DPC) and a k-means clustering algorithm, provides a self-adaptive k-means clustering algorithm, self-adaptively determines the number of targets to be positioned based on extracted characteristic points, and clusters characteristic point sets of different targets; coarse matching is carried out through a nearest neighbor ratio algorithm, optimal geometric constraint is constructed through voting of feature points, fine matching is carried out, and multi-target accurate positioning is achieved. The method can accurately position targets in complex environments such as rotation, scale transformation, partial shielding, illumination transformation and the like for different types and numbers of targets to be positioned, and has better robustness.
Description
Technical Field
The invention relates to the technical field of computer vision and image processing, in particular to a multi-target positioning method based on self-adaptive k-means clustering.
Background
Image matching is an important content of computer vision and pattern recognition, and is widely applied in the fields of image registration, image stitching, three-dimensional reconstruction and the like. Currently, image matching algorithms are mainly classified into two categories, gray-level-based matching and feature-based matching. The matching algorithm based on the gray level matches through the regional attribute in the image sampling window, the matching precision is high, but the matching algorithm is easy to be influenced by the environment and sensitive to the change of the gray level of the image; based on a feature matching algorithm, feature description is carried out by detecting stable features in an image and utilizing neighborhood pixel information, matching is completed according to the similarity of the calculated feature descriptors, the algorithm has strong anti-interference capability, high matching speed and high robustness. Therefore, scholars have proposed many excellent feature-based matching algorithms. Lowe proposes a SIFT algorithm which can adapt to rotation and scale scaling transformation and is insensitive to illumination change, but because of the existence of texture similar areas in images, mismatching is easily caused by only adopting neighborhood information as descriptors, and the application of the algorithm is limited to a certain extent. Jiao Yang et al combine k-means with SIFT algorithm, utilize k-means algorithm to cluster eigenvector matrix, have raised the matching rate of the picture; zhang Ligong et al firstly perform logarithmic transformation on the image, and then perform iterative matching on the image characteristics to realize multi-objective matching of sonar images; dong Jintao et al only use the characteristic point of the single-layer Gaussian pyramid to carry on the rough matching in the matching process, combine GMS algorithm and RANSAC algorithm to calculate affine transformation matrix at the same time, improve the real-time of the image matching; wang Ting et al propose that geometric constraints are combined with SIFT algorithm to match images, epipolar constraints are added in the matching stage, and good affine invariance is achieved; li Yungong et al utilize iterative least squares fitting to construct a functional model to reject mismatching points, which has advantages in matching time and accuracy.
Although the improved SIFT algorithm improves the matching efficiency, in the multi-target positioning process, the SIFT algorithm cannot accurately position all targets, the number of targets to be positioned must be determined, and different target feature points are clustered.
Disclosure of Invention
Aiming at the problems, the invention provides a multi-target positioning algorithm based on self-adaptive k-means clustering, which comprises the following specific steps:
1. detecting a feature point set of an image to be matched and calculating a feature point main direction
(1) Convolving the image with Gaussian functions of different scale factors to construct a Gaussian scale space, and subtracting images in adjacent scale spaces to construct a Gaussian differential scale space;
(2) Calculating a local extremum point of the Gaussian difference scale space as a characteristic point;
(3) Calculating the gradient size and direction of pixels in the neighborhood taking the feature point as the center, and counting the gradient size and direction of all pixels to generate a gradient direction histogram, wherein the maximum value direction of the histogram is the main direction of the feature point;
(4) Dividing the sampling window into a plurality of subareas, counting gradient information in the subareas, constructing feature vectors, and carrying out normalization processing on the feature vectors in order to further reduce the influence of illumination change.
2. The number of clusters and the initial cluster center are determined based on a density peak clustering algorithm (DPC).
(1) Calculating the cut-off distance of the characteristic points and the local density and distance of each characteristic point;
(2) And calculating the gamma value of each characteristic point, arranging the gamma values in a descending order, and selecting the first k data as an initial clustering center. The calculation formula of gamma is as follows:
γ i ＝ρ i *δ i
rho in the above i Is of local density delta i Is the distance.
3. And (3) carrying out iterative clustering on the clustering centers by using a k-means clustering algorithm to obtain a final clustering result. And calculating the distance between all the feature points and the clustering center, classifying the data to be classified according to the nearest neighbor principle, calculating the average value of the coordinates of all the feature points in each cluster after classifying all the feature points, taking the average value of the coordinates as a new clustering center, continuously and iteratively determining to finally obtain k clustering centers, and clustering the rest feature points to the cluster where the feature points are closest in distance and have higher local density than the rest feature points.
4. Performing rough matching of the characteristic points by utilizing a nearest neighbor ratio algorithm, and performing rough elimination on the characteristic points;
searching feature points closest to and next closest to the template image feature points in the feature point set of the image to be matched, and calculating the closest Euclidean distance dis nt And next closest Euclidean distance dis snt If the ratio of (2) is satisfied:
and if the characteristic points are successfully matched, otherwise, eliminating the characteristic point pair, wherein T is a set threshold value.
5. And constructing optimal geometric constraint by utilizing the characteristic point voting to carry out fine matching.
According to the descending order of the matching confidence degree of the feature point pairs in the rough matching, 3 pairs of feature points are selected randomly from the first n pairs of matching feature points through iteration to vote, and 3 pairs of feature point pairs with the best fitness are selected to establish a local coordinate system. Based on the constructed local coordinate system, constructing a straight line primitive by the feature points and the coordinate origin, expressing vector coordinates, calculating the coordinate similarity of the vector coordinates in the local coordinate system, comparing the similarity with a set threshold value, and eliminating mismatching points.
(1) The construction method of the middle local coordinate system comprises the following steps:
feature point matching confidence degree descending order based on coarse matching is used for selecting top n pairs of feature points to form a set R= { (M) i ,N i ) I=1, …, n }, where M i And N i The method is a pair of matched characteristic points, 3 characteristic points are selected from a set R, 3 characteristic points are selected from the set R to construct a local coordinate system, the similarity of the rest characteristic points in the set in the local coordinate system is utilized to conduct voting score, and finally 3 characteristic points with highest scores are selected to construct the local coordinate system.
The feature point similarity calculation formula is as follows:
in the formula { (L) ji ,L ji ' i=1, 2,3} is the matching point pair M j And N j Euclidean distance to 3 matching feature points.
(2) The coordinate similarity of the feature point straight line primitive is calculated as follows:
(1) the linear primitive vectors are represented by coordinates, and form a coordinate set: Ω= { p i (x,y),q i (x,y)|i＝1,…,N}
The linear primitive vector represents the calculation formula by coordinates:
where k is the slope of a straight line, (x) 1 ,y 1 ) Is the coordinates of the origin of the straight primitive, (x) 2 ,y 2 ) Is sitting at the endAnd (5) marking.
(2) The coordinate conversion formula for converting vector coordinates into a local coordinate system is as follows:
where α, β are the coordinates of the vectorized straight line primitive in the local coordinate system.
(3) Calculating coordinate similarity of straight line primitive
Wherein P is M And P N Is the local coordinates of the straight line primitive.
6. And calculating a transformation model by using the correct matching points to realize multi-objective accurate matching.
The invention has the following advantages: (1) The DPC algorithm is combined with the k-means, the clustering number and the clustering center are determined through the DPC algorithm, and then the initial characteristic points are clustered through the k-means algorithm, so that the characteristic point set can be effectively separated. (2) The algorithm of the invention realizes the accurate and rapid matching of multiple targets, can still realize the rapid matching of the targets when the targets to be matched are subjected to rotation transformation, scale transformation, illumination influence, shielding and other factors, has scale and rotation invariance, and has higher stability when the matching effect is not influenced by shielding and illumination change.
Drawings
Fig. 1 is an algorithm flow chart.
Fig. 2 is a schematic diagram of local coordinate calculation of a straight line primitive.
FIG. 3 is a graph of matching results for strip workpieces.
Figure 4 is a graph of the results of the L form of the workpiece matching.
Detailed Description
1. Inputting an image to be matched, obtaining characteristic points of the image to be matched and calculating main directions of the characteristic points
(1) The method comprises the steps that an image to be matched is convolved with Gaussian functions of different scale factors to construct a Gaussian scale space, and in order to improve stability and accuracy of extreme point detection, a Gaussian differential scale space is constructed by subtracting images in adjacent scale spaces in the Gaussian scale space;
(2) And calculating the local extremum point of the Gaussian difference scale space as a characteristic point. Comparing the to-be-detected point with 26 pixels in the neighborhood of 8 pixels in the same scale and the neighborhood of 9 multiplied by 2 pixels in the corresponding position of the adjacent scale, wherein if the gray value of the to-be-detected pixel is completely smaller or larger than the gray value of 26 pixels in the neighborhood, the pixel is a local extremum point in the Gaussian scale space, and the local extremum point set is a characteristic point set
(3) In order to ensure that the descriptor has rotation invariance, a reference direction is required to be allocated to the feature points, the gradient sizes and directions of pixels in a neighborhood taking the feature points as the center are calculated, gradient directions of all pixels are counted to generate a gradient direction histogram, and the maximum direction of the histogram is the main direction of the feature points;
(4) Since the coordinate axes do not coincide with the main direction, the sampling window needs to be rotated to the main direction of the feature point. Dividing the sampling window into 4×4 sub-regions, counting gradient information in 8 directions in the sub-regions, constructing 128-dimensional feature vectors, and carrying out normalization processing on the feature vectors in order to further reduce the influence of illumination variation.
2. The number of clusters K and the initial cluster center are determined based on a density peak clustering algorithm (DPC).
During the multi-target matching process, the feature points of different targets in the feature point set must be clustered, so that interference of other target feature points in the matching process is avoided. The DPC algorithm can determine the clustering centers and the clustering number by using density peaks in the sample data, and solves the problem of selecting the initial clustering centers and the clustering number. The specific method comprises the following steps:
(1) Calculating a feature point set of an image to be matchedIs defined by the local density ρ of each feature point i And distance delta i ；
Any one feature point has two indexes: local density ρ i And distance delta i To determine the cluster center, both local density and distance influencing factors are considered.
(2) The gamma value of each feature point is calculated, and the calculation formula of gamma is as follows:
γ i ＝ρ i *δ i
the larger the gamma value is, the higher the probability that the feature point becomes a cluster center is, so that the feature point is arranged in a descending order, and the first k data are selected as initial cluster centers.
3. Iterative clustering is carried out on the clustering centers by using a k-means clustering algorithm, so that a final clustering result is obtained, and a feature point set is effectively separated;
based on the initial clustering centers, calculating the distances from all the feature points to the clustering centers, classifying the feature points to be classified according to the nearest neighbor principle, calculating the average value of the coordinates of the feature points of each category after classifying all the feature points, taking the average value as a new clustering center, and continuously and iteratively determining to finally obtain k clustering centersWherein->And clustering the rest characteristic points to the class cluster which is closest to the rest characteristic points and has local density higher than that of the rest characteristic points.
4. Performing rough matching of the characteristic points by utilizing a nearest neighbor ratio algorithm, and performing rough elimination on the characteristic points;
searching feature points closest to and next closest to the template image feature points in the feature point set of the image to be matched, and calculating the closest Euclidean distance dis nt And next closest Euclidean distance dis snt If the ratio of (2) is satisfied:
and if the characteristic points are successfully matched, otherwise, eliminating the characteristic point pair, wherein T is a set threshold value.
5. And constructing optimal geometric constraint by utilizing the characteristic point voting to carry out fine matching.
According to the descending order of the matching confidence degree of the feature point pairs in the rough matching, 3 pairs of feature points are selected randomly from the first n pairs of matching feature points through iteration to vote, and 3 pairs of feature point pairs with the best fitness are selected to establish a local coordinate system. Based on the constructed local coordinate system, constructing a straight line primitive by the feature points and the coordinate origin, expressing vector coordinates, calculating the coordinate similarity of the vector coordinates in the local coordinate system, comparing the similarity with a set threshold value, and eliminating mismatching points.
(1) Construction of local coordinate System
Feature point matching confidence degree descending order based on coarse matching is used for selecting top n pairs of feature points to form a set R= { (M) i ,N i ) I=1, …, n }, where M i And N i The method is a pair of matched characteristic points, 3 characteristic points are selected from a set R to construct a local coordinate system, the similarity of the rest characteristic points in the set in the local coordinate system is utilized to conduct voting score, and finally 3 characteristic points with highest scores are selected to construct the local coordinate system.
The feature point similarity calculation formula is as follows:
in the formula { (L) ji ,L ji ' i=1, 2,3} is the matching point pair M j And N j Euclidean distance to 3 matching feature points.
(2) Calculating vectorized coordinates of feature point straight line elements
Assume that the extracted 3 feature point pairs are { (M) j ,N j ) I j=1, 2,3}, M in each of the two images 1 And N 1 For origin of coordinates, connect M 1 M 2 ,M 1 M 3 And N 1 N 2 ,N 1 N 3 Respectively establishing a coordinate system O-XY and O-X 'Y', and utilizing straight line elementsVectorized coordinates are used as base information P of a coordinate system O-XY A (P ax ,P ay ) And P B (P bx ,P by ) The base information of the coordinate system O-X 'Y' is calculated in the same manner as shown in fig. 2.
The linear primitive vectors are represented by coordinates, and form a coordinate set: Ω= { p i (x,y),q i (x,y)|i＝1,…,N}
The linear primitive vector represents the calculation formula by coordinates:
where k is the slope of a straight line, (x) 1 ,y 1 ) Is the coordinates of the origin of the straight primitive, (x) 2 ,y 2 ) Is the end point coordinates.
(3) Similarity calculation
To calculate the similarity of the straight line primitive, its vector coordinates are converted into coordinates in the local coordinate system, and the conversion formula is as follows:
where α, β are the coordinates of the vectorized straight line primitive in the local coordinate system.
The coordinate similarity of the straight line primitive is calculated as follows:
wherein P is M And P N Is the local coordinates of the straight line primitive.
By comparing the similarity S with a set threshold t s Comparing if S is less than t s The feature point pairs are correctly matched; otherwise, the matching point is a mismatching point; multiple matching experiments prove that when t s When 0.03 is taken, the matching effect is optimal.
6. And calculating a transformation model by using the correct matching points to realize multi-objective accurate matching.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present disclosure describes embodiments, not every embodiment is provided with a separate embodiment, and that this description is provided for clarity only, and that the disclosure is not limited to the embodiments described in detail below, and that the embodiments described in the examples may be combined as appropriate to form other embodiments that will be apparent to those skilled in the art.
Claims (1)
1. A multi-target positioning method based on self-adaptive k-means clustering is characterized by comprising the following steps: the method comprises the following steps:
s1, detecting a feature point set of an image to be matched and calculating a feature point main direction;
s2, determining the number of clusters and an initial cluster center based on a density peak clustering algorithm;
s3, performing iterative clustering on the clustering centers by using a k-means clustering algorithm to obtain a final clustering result;
s4, performing rough matching of the characteristic points by utilizing a nearest neighbor ratio algorithm, and performing rough elimination of the characteristic points;
s5, constructing an optimal geometric constraint by utilizing the characteristic point voting to carry out fine matching;
s6, calculating a transformation model by using correct matching points to realize multi-objective accurate matching;
wherein S1 comprises:
(1) Convolving the image with Gaussian functions of different scale factors to construct a Gaussian scale space, and subtracting images in adjacent scale spaces to construct a Gaussian differential scale space;
(2) Calculating a local extremum point of the Gaussian difference scale space as a characteristic point;
(3) Calculating the gradient size and direction of pixels in the neighborhood taking the feature point as the center, and counting the gradient size and direction of all pixels to generate a gradient direction histogram, wherein the maximum value direction of the histogram is the main direction of the feature point;
(4) Dividing the sampling window into a plurality of subareas, counting gradient information in the subareas, constructing feature vectors, carrying out normalization processing on the feature vectors in order to further reduce the influence of illumination variation,
s2 comprises the following steps:
(1) Calculating the cut-off distance of the characteristic points and the local density and distance of each characteristic point;
(2) The gamma value of each characteristic point is calculated, the gamma values are arranged in a descending order, and the first k data are selected as the calculation formula of the initial clustering center gamma as follows:
γ i ＝ρ i *δ i
rho in the above i Is of local density delta i In order to be a distance from each other,
in the step S3, the processing unit,
calculating the distance between all feature points and the clustering center, classifying the data to be classified according to the nearest neighbor principle, calculating the average value of the coordinates of all feature points in each cluster after classifying all feature points, taking the average value of the coordinates as a new clustering center, continuously and iteratively determining to finally obtain k clustering centers, clustering the rest feature points to the cluster where the feature points are nearest and the local density is higher than that of the rest feature points,
in the step S3, the processing unit,
searching feature points closest to and next closest to the template image feature points in the feature point set of the image to be matched, and calculating the closest Euclidean distance dis nt And next closest Euclidean distance dis snt If the ratio of (2) is satisfied:
the feature point matching is successful, otherwise, the feature point pair is eliminated, T is a set threshold,
in S5, the processing unit is configured to,
according to the descending order of the matching confidence of the feature point pairs in the rough matching, 3 pairs of feature points are randomly selected from the first n pairs of matching feature points through iteration to vote, 3 pairs of feature point pairs with the best fitness are selected to establish a local coordinate system, a straight line primitive is established based on the established local coordinate system, the feature points and the coordinate origin are established, vector coordinates are expressed, the coordinate similarity of the vector coordinates in the local coordinate system is calculated, the similarity is compared with a set threshold value, the mismatching point is removed,
the construction method of the local coordinate system comprises the following steps:
feature point matching confidence degree descending order based on coarse matching is used for selecting top n pairs of feature points to form a set R= { (M) i ,N i ) I=1, …, n }, where M i And N i Selecting 3 feature points from the set R, constructing a local coordinate system, voting the similarity of the rest feature points in the set in the local coordinate system, and finally selecting the 3 feature points with the highest scores to construct the local coordinate system;
the feature point similarity calculation formula is as follows:
in the formula { (L) ji ,L ji ' i=1, 2,3} is the matching point pair M j And N j Euclidean distance to 3 matching feature points;
the coordinate similarity of the feature point straight line primitive is calculated as follows:
(1) the linear primitive vectors are represented by coordinates, and form a coordinate set: Ω= { p i (x,y),q i (x,y)|i＝1,L,N}
The linear primitive vector represents the calculation formula by coordinates:
where k is the slope of a straight line, (x) 1 ,y 1 ) Is the coordinates of the origin of the straight primitive, (x) 2 ,y 2 ) Is the end point coordinates;
(2) the coordinate conversion formula for converting vector coordinates into a local coordinate system is as follows:
wherein alpha and beta are coordinates of vectorized straight line elements in a local coordinate system;
(3) calculating coordinate similarity of straight line primitive
Wherein P is M And P N Is the local coordinates of the straight line primitive.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111634014.5A CN114358166B (en) | 2021-12-29 | 2021-12-29 | Multi-target positioning method based on self-adaptive k-means clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111634014.5A CN114358166B (en) | 2021-12-29 | 2021-12-29 | Multi-target positioning method based on self-adaptive k-means clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114358166A CN114358166A (en) | 2022-04-15 |
| CN114358166B true CN114358166B (en) | 2023-11-07 |
Family
ID=81103034
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111634014.5A Active CN114358166B (en) | 2021-12-29 | 2021-12-29 | Multi-target positioning method based on self-adaptive k-means clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114358166B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385480B (en) * | 2023-02-03 | 2023-10-20 | 腾晖科技建筑智能(深圳)有限公司 | Detection method and system for moving object below tower crane |
| CN116129325A (en) * | 2023-02-20 | 2023-05-16 | 城云科技（中国）有限公司 | Urban treatment image target extraction method and device and application thereof |
| CN116580218A (en) * | 2023-05-17 | 2023-08-11 | 中国第一汽车股份有限公司 | Control method, device, equipment and storage medium |
| CN116796214B (en) * | 2023-06-07 | 2024-01-30 | 南京北极光生物科技有限公司 | Data clustering method based on differential features |
| CN118094274B (en) * | 2024-02-27 | 2025-03-14 | 国网山西省电力公司电力科学研究院 | Method and device for identifying full-level topology of low-voltage distribution area |
| CN118968110B (en) * | 2024-10-15 | 2025-05-16 | 山东大学 | Feature matching method, system, medium, product and equipment based on K clustering |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103150737A (en) * | 2013-01-18 | 2013-06-12 | 西北工业大学 | Real-time space target feature point tracking method suitable for space tethered robot |
| WO2014067296A1 (en) * | 2012-11-05 | 2014-05-08 | 深圳市恩普电子技术有限公司 | Method for identifying, tracing and measuring internal and external membranes of vessel |
| CN104574347A (en) * | 2013-10-24 | 2015-04-29 | 南京理工大学 | On-orbit satellite image geometric positioning accuracy evaluation method on basis of multi-source remote sensing data |
| CN108154118A (en) * | 2017-12-25 | 2018-06-12 | 北京航空航天大学 | A kind of target detection system and method based on adaptive combined filter with multistage detection |
| WO2019245320A1 (en) * | 2018-06-22 | 2019-12-26 | 삼성전자주식회사 | Mobile robot device for correcting position by fusing image sensor and plurality of geomagnetic sensors, and control method |
| CN113177463A (en) * | 2021-04-27 | 2021-07-27 | 中国工商银行股份有限公司 | A kind of target positioning method and device in mobile scene |
-
2021
- 2021-12-29 CN CN202111634014.5A patent/CN114358166B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014067296A1 (en) * | 2012-11-05 | 2014-05-08 | 深圳市恩普电子技术有限公司 | Method for identifying, tracing and measuring internal and external membranes of vessel |
| CN103150737A (en) * | 2013-01-18 | 2013-06-12 | 西北工业大学 | Real-time space target feature point tracking method suitable for space tethered robot |
| CN104574347A (en) * | 2013-10-24 | 2015-04-29 | 南京理工大学 | On-orbit satellite image geometric positioning accuracy evaluation method on basis of multi-source remote sensing data |
| CN108154118A (en) * | 2017-12-25 | 2018-06-12 | 北京航空航天大学 | A kind of target detection system and method based on adaptive combined filter with multistage detection |
| WO2019245320A1 (en) * | 2018-06-22 | 2019-12-26 | 삼성전자주식회사 | Mobile robot device for correcting position by fusing image sensor and plurality of geomagnetic sensors, and control method |
| CN113177463A (en) * | 2021-04-27 | 2021-07-27 | 中国工商银行股份有限公司 | A kind of target positioning method and device in mobile scene |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114358166A (en) | 2022-04-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114358166B (en) | Multi-target positioning method based on self-adaptive k-means clustering | |
| CN110097093B (en) | Method for accurately matching heterogeneous images | |
| CN107067415B (en) | A kind of object localization method based on images match | |
| CN106355577B (en) | Fast Image Matching Method and System Based on Feature State and Global Consistency | |
| CN106919944B (en) | ORB algorithm-based large-view-angle image rapid identification method | |
| CN102147858B (en) | License plate character identification method | |
| CN110930456A (en) | Three-dimensional identification and positioning method of sheet metal part based on PCL point cloud library | |
| CN108491838B (en) | Reading Method of Pointer Type Instrument Based on SIFT and HOUGH | |
| CN111768447A (en) | Monocular camera object pose estimation method and system based on template matching | |
| CN108182705A (en) | A kind of three-dimensional coordinate localization method based on machine vision | |
| CN112364881A (en) | Advanced sampling consistency image matching algorithm | |
| CN106570475B (en) | A kind of dark-red enameled pottery seal search method | |
| CN102446356A (en) | Parallel self-adaptive matching method for obtaining remote sensing images with uniformly distributed matching points | |
| CN113313701A (en) | Electric vehicle charging port two-stage visual detection positioning method based on shape prior | |
| CN110942473A (en) | Moving target tracking detection method based on characteristic point gridding matching | |
| CN111914892A (en) | Recognition method of vehicle type and vehicle logo based on tire detection | |
| CN110246165A (en) | It improves visible images and SAR image matches the method and system of Quasi velosity | |
| CN117079272B (en) | A method for recognizing bullet bottom dimple marks by combining manual features and learning features | |
| CN118823126A (en) | A method for positioning PCB solder-free optical devices | |
| CN115641367A (en) | A Method of Infrared and Visible Image Registration Based on Multi-Level Feature Matching | |
| CN111199558A (en) | Image matching method based on deep learning | |
| CN113570647A (en) | A spatial registration method of stereo objects between oblique photography and remote sensing optical images | |
| CN111709426A (en) | Diatom identification method based on contour and texture | |
| CN114170420A (en) | Multi-spot image gravity center positioning method for micro-lens array imaging system | |
| Han et al. | Accurate and robust vanishing point detection method in unstructured road scenes |
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 |