CN109241870B - Coal mine underground personnel identity identification method based on gait identification - Google Patents
Coal mine underground personnel identity identification method based on gait identification Download PDFInfo
- Publication number
- CN109241870B CN109241870B CN201810936703.3A CN201810936703A CN109241870B CN 109241870 B CN109241870 B CN 109241870B CN 201810936703 A CN201810936703 A CN 201810936703A CN 109241870 B CN109241870 B CN 109241870B
- Authority
- CN
- China
- Prior art keywords
- gait
- matrix
- coal mine
- image
- identity
- 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.)
- Expired - Fee Related
Links
- 230000005021 gait Effects 0.000 title claims abstract description 99
- 239000003245 coal Substances 0.000 title claims abstract description 39
- 238000000034 method Methods 0.000 title claims abstract description 36
- 239000011159 matrix material Substances 0.000 claims abstract description 60
- 239000013598 vector Substances 0.000 claims abstract description 21
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 230000009467 reduction Effects 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 238000000605 extraction Methods 0.000 claims abstract description 4
- 238000005259 measurement Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 10
- 238000001914 filtration Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000000737 periodic effect Effects 0.000 claims description 3
- 238000011410 subtraction method Methods 0.000 claims description 3
- 230000002194 synthesizing effect Effects 0.000 claims description 3
- 238000000691 measurement method Methods 0.000 claims description 2
- 238000012544 monitoring process Methods 0.000 abstract description 9
- 238000005286 illumination Methods 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000002817 coal dust Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000036544 posture Effects 0.000 description 1
- 238000000547 structure data Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
- G06V40/25—Recognition of walking or running movements, e.g. gait recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a gait recognition-based method for identifying the identity of a coal mine underground person, which comprises the steps of preprocessing an acquired coal mine underground gait image to obtain a high-dimensional gait image vector, setting a neighborhood weight matrix by using a self-adaptive measurement mode, constructing an objective function, optimizing the objective function, projecting test set data to a low-dimensional feature classification space by using an optimal projection matrix, and realizing the dimensionality reduction and gait feature extraction of a high-dimensional space data sample point; and during identification, inputting the gait image sequence to be identified into the K-NN classifier to realize classification, identification and monitoring of the gait images. The method is less affected by illumination, can better judge the local information and the category information of the subspace where each high-dimensional data point is located, has high recognition rate and good stability, and can provide reliable information for automatic monitoring of the position of underground personnel in a coal mine, automatic identification of the identity of the personnel and the like.
Description
Technical Field
The invention relates to a method for identifying the identity of a person in a coal mine well based on gait recognition, which relates to the fields of image recognition, communication and the like.
Background
Coal is the most important basic energy in China and is one of strategic development energy in China, and the energy structure mainly based on coal in China is not changed for a long time. With the continuous development of economy, the total coal demand is continuously increased, and coal mine safety production accidents therewith are also gradually increased, so that the coal mine safety problem needs to be solved urgently. As one of six systems for coal mine underground safety risk avoidance, a coal mine underground personnel position monitoring system is used for monitoring underground personnel positions and has the functions of managing and inquiring the time of taking in and out a well by a card-carrying personnel, the time of taking in and out a key area, the time of limiting the time of taking in and out an area and the like. The identity recognition of personnel under the mine is a key link of the operation of the system, the widely used biometric identification technologies at present comprise fingerprint identification, face identification, iris identification, palm shape identification and the like, and the identity recognition technology of personnel under the mine usually adopts fingerprint identification and face identification. Although the biological recognition technology is more mature, the recognition rate is higher under normal environment. However, the underground environment is severe, the illuminance is low, the space is limited, the air is humid, the content of floating impurities such as coal dust in a roadway is high, and the like, so that fingerprints and human faces are blurred, and the identification rate of the identities of underground personnel through fingerprint identification and human face identification is greatly reduced.
Gait recognition aims to identify according to the body type characteristics and walking postures of people. As a new biological feature recognition technology, the technology rapidly draws the attention of researchers at home and abroad due to the recognition potential under the conditions of long distance and low video quality. Compared with other biological feature identification technologies such as human faces, fingerprints and the like, the gait identification technology has the advantages of being long in distance, uncontrolled, difficult to disguise, small in environmental influence and the like, and can still identify and sense in severe environments under mines.
At present, various gait feature extraction methods are available, including Hough transformation, PCA, wavelet transformation, time-frequency domain analysis, Trace transformation, linear interpolation, tensor discriminant analysis and the like, but the methods are all linear methods, so that the method is effective in processing linear structure data, and a nonlinear essential structure of a gait image cannot be obtained. In addition, although the classical LPP algorithm projects in the subspace, the distance characteristics of the data in the initial space are still used to perform the neighbor selection and the similarity determination. Since the original data includes various redundant features (noise, etc.), the similarity between the data cannot be truly reflected. Because the gait image sequence is characterized by high dimension, complexity and changeability and is nonlinear data, the recognition rate of the methods for gait recognition is not high and the application is limited.
Disclosure of Invention
The invention provides a method for identifying the identity of underground coal mine personnel based on gait recognition, which adopts a self-adaptive mode to select a nearest neighbor point, constructs an objective function according to different constraints between similar and heterogeneous samples, and obtains an optimal projection matrix by minimizing the objective function, thereby completing the classified recognition and monitoring of underground coal mine operators by projection dimension reduction of high-dimensional gait data sample points. The monitoring method is less affected by illumination, and can better judge the local information and the category information of the subspace where each high-dimensional data point is located, so that the method has higher identification accuracy and identification stability, and can provide reliable information for automatic monitoring of the position of underground personnel in a coal mine, automatic identification of personnel identity and the like. The method solves the problems that the recognition rate of the conventional gait recognition method for gait recognition is not high and the application is limited.
The invention provides a method for identifying the identity of underground coal mine personnel based on gait recognition, which comprises the following steps:
A. acquiring gait video sequences of a plurality of operators in a coal mine to form a gait image database;
B. preprocessing gait video sequences in a gait image database to obtain gait video image vectors of a high-dimensional space, and then selecting one half of the gait video sequences of each operator as a training set and the other half of the gait video sequences as a test set;
C. carrying out dimensionality reduction on the gait video image vectors in the training set, and solving an optimal projection matrix;
D. using the optimal projection matrix to project the gait video image data of the test set to a low-dimensional gait image data feature classification space to realize the extraction of the gait feature data;
E. inputting the gait image sequence to be identified into a K-NN classifier, performing classification identification on the test set by using the K-NN classifier according to the extracted gait features, and identifying the identity of the underground coal mine personnel according to the classification identification result.
Further, in step a, a gait video image sequence of the operator is collected by the camera, and the operator walks at 3 viewing angles of the front, the incline and the side relative to the camera respectively, and each person collects 4 color image sequences at each viewing angle.
Further, in step B, the method for preprocessing the gait video sequence comprises: firstly, extracting a single frame image from an original gait video image collected in the step A to perform gray level transformation, then performing median filtering, calculating the median of each pixel point of the gait video sequence after the median filtering frame by frame to be used as a background image of the gait video sequence, extracting a human body target by using a background subtraction method, selecting the gait outline of the human body target by using a rectangular frame with normalized scale, intercepting the gait outline, normalizing the gait outline into a size of 128 multiplied by 64 pixels according to a 1:1 scaling ratio, synthesizing a periodic gait image into a binary image by using a detected gait period, and then converting each image matrix into vector representation to obtain a high-dimensional gait video image vector.
Further, in step B, the aspect ratio of the rectangular frame is set to 128: 64.
Further, in step C, the solution of the optimal projection matrix is specifically as follows:
(1) performing dimensionality reduction on the gait video image vector of the high-dimensional space in the training set;
(2) constructing a weighted neighborhood graph according to a k-nearest neighbor criterion, and setting a neighborhood weight matrix by using a self-adaptive nearest neighbor measurement mode;
(3) defining a target function according to the neighborhood weight matrix, wherein the target function is used for ensuring that the type of the same or different nearest neighbor sample points is unchanged after projection, the distance of the same type of points is reduced after projection, and the distance of the different types of points is enlarged after projection;
(4) and optimizing the objective function, and solving the optimized objective function to obtain an optimal projection matrix.
Further, let n sample point sets X ═ X of gait video image vectors in high-dimensional space1,x2,,xnD, dimension D; the sample point set Y of the gait video image vector of the corresponding low-dimensional space is { Y ═ Y1,y2,,ynD, dimension;
the adaptive nearest neighbor measurement method comprises the following steps: d (x)i,xj)=(xi-xj)TΣ*(xi-xj)
Wherein: w ═ W-1BW-1=W-1/2(W-1/2BW-1/2)W-1/2=W-1/2B*W-1/2
Σ*=W-1/2(W-1/2BW-1/2+δE)W-1/2=W-1/2(B*+δE)W-1/2
Sigma is the covariance matrix, delta is the softening parameter, [ sigma ]*Is a defined covariance matrix, W is the summed intra-class covariance matrix, B is the inter-class covariance matrix, and the element values in W and B are computed from the k nearest neighbors around each point, each point having its corresponding element value in the W matrix and element value in the B matrix.
Further, in step (2), the neighborhood weight matrix H ═ Hij},
In the formula, N (x)i)、N(xj) Are respectively a sample point xi、xjK sets of nearest neighbor points, ciAnd cjIs a sample point xiAnd xjAnd gamma is a control parameter.
Further, the objective function is:
further, the objective function is optimized as:
d is a diagonal matrix, D ═ Djj},The matrix obtained by D-H is a Laplacian matrix, namely L' ═ D-H, and A is an optimal projection matrix.
Further, an optimal projection matrix A is obtained by decomposing the eigenvalues, and k minimum non-zero eigenvalues λ are obtained assuming that k is the dimensionality of the sample point reduction1,λ2,…,λkRespectively corresponding to the feature vector a1,a2,…,akObtaining the optimal projection matrix A ═ a meeting the optimized objective function1,a2,,ak}。
The invention has the beneficial effects that: the method is less affected by illumination, can better judge the local information and the category information of the subspace where each high-dimensional data point is located, has high recognition rate and good stability, and can provide reliable information for automatic monitoring of the position of the underground coal mine personnel, automatic identification of the identity of the personnel and the like.
Drawings
Fig. 1 is a schematic diagram of a person identification process according to the present invention.
Detailed Description
The invention is further described with reference to the following figures and specific examples.
The identity recognition method of the underground coal mine personnel based on gait recognition can accurately recognize and monitor the identity of the underground coal mine personnel in the first time, improve the accuracy and stability of the underground coal mine gait recognition, and provide a guarantee for the safety operation of the underground coal mine personnel.
The personnel identity recognition method of the invention is shown in figure 1:
1, (101) acquiring gait video sequences of a plurality of underground operators from an underground coal mine roadway through a camera to form a gait database.
And acquiring the gait video sequence of each underground operator at different visual angles, and detecting the gait cycle of the gait video sequence at each visual angle.
In this example, gait video sequences of 20 persons were collected, each person walking at 3 views (0 °,45 °,90 ° from the image plane, respectively) in front, oblique and lateral directions with respect to the camera, and each person collected 4 sequences per view, for a total of 240 sequences. A sequence of color images of 352 x 240 size, with an average length of about 100 frames, is taken at a rate of 25 frames per second.
And 2, (102) preprocessing the underground coal mine gait video sequence collected by the camera, and dividing the underground coal mine gait video sequence into a training set and a testing set. Wherein the pretreatment process comprises the following steps:
and (4) extracting a single-frame color image from the original gait video image collected in the step (101) to perform gray level transformation, and then performing median filtering. And calculating the median of each pixel point frame by frame of the gait video sequence after median filtering processing, using the median as a background image of the gait video sequence, and extracting a human body target by using a background subtraction method. Selecting the gait contour of the human body target by using a rectangular frame with normalized scale, intercepting the gait contour, normalizing the gait contour into the size of 128 multiplied by 64 pixels according to the scaling ratio of 1:1, wherein the aspect ratio of the rectangular frame is preferably as follows: 128:64. Synthesizing a periodic gait image into a binary image by using the detected gait cycle, and converting each image matrix into vector representation by using Matlab, wherein the ith gait image can be represented as a vector Xi={x1,x2,,xdI-1, 2, d denotes the dimension of the vector.
After the gait video sequence is preprocessed, half of the gait video sequence is selected from the gait images of each person in the gait database to be used as a training set, and the remaining half is used as a testing set.
(103) carrying out dimensionality reduction on the vector converted by the gait image in the training set, wherein the dimensionality reduction is as follows:
suppose that n high-dimensional data sample point sets X before dimensionality reduction are { X ═ X1,x2,,xnD in dimension, and after dimension reduction projection, corresponding low-dimensional data sample point set Y is { Y ═ Y1,y2,,ynD, dimension.
Defining a new adaptive nearest neighbor metric: d (x)i,xj)=(xi-xj)TΣ*(xi-xj) Wherein:
Σ=W-1BW-1=W-1/2(W-1/2BW-1/2)W-1/2=W-1/2B*W-1/2
Σ*=W-1/2(W-1/2BW-1/2+δE)W-1/2=W-1/2(B*+δE)W-1/2
sigma is a covariance matrix, and to avoid infinite value of Sigma in the measurement, a softening parameter delta is introduced to be defined as Sigma*,Σ*Is a limited covariance matrix, and takes δ to be 1, W is the summed intra-class covariance matrix, B is the inter-class covariance matrix, and the element values in the W matrix and the B matrix are adaptively determined by the calculation of k nearest neighbors around each point, each point has its own corresponding element value in the W matrix and element value in the B matrix.
(104) constructing a weighted neighborhood graph G according to a k-nearest neighbor criterion, wherein G is an undirected graph. Let N (x)i)、N(xj) Are respectively a sample point xi、xjK nearest neighbor point sets, and the neighbor weight matrix H ═ Hij}。
In the formula, ciAnd cjIs a sample point xiAnd xjA class C tag of (1); gamma is a control parameter.
(105) defining the objective function as:the objective function aims to ensure that the type of the same or different nearest neighbor sample points is unchanged after projection, the distance of the same type of points after projection is reduced, and the distance of the different types of points after projection is enlarged.
(106) adding a constraint ATXDXTE, optimizing the objective function as:
in the formula: d is a diagonal matrix, D ═ Djj},Thus, the matrix obtained by D-H is a Laplacian matrix, i.e., L' ═ D-H.
And 7, (107) converting the objective function minimization problem into a solution matrix eigenvalue and eigenvector problem, wherein the optimal projection matrix A can be obtained by decomposing the eigenvalues.
Assuming that k is the reduced dimension of the sample point, the k minimum non-zero eigenvalues λ are found1,λ2,…,λkRespectively corresponding to the feature vector a1,a2,…,akObtaining the optimal projection matrix A ═ a satisfying the formula1,a2,,ak}。
In order to select parameters, an optimal softening parameter delta and the number k of neighbors are selected, 5-fold cross validation is carried out on training samples, based on a Grid Search method (Grid Search), the Search range of delta is {0,0.01,0.1,0.2,0.5,1,2,5}, the Search range of k is [1,50], the step length is 1, and the parameter combination with the highest gait recognition accuracy is searched to serve as the optimal parameters of the model.
And 8, (108) projecting the data of the test set to a low-dimensional data feature classification space according to the optimal projection matrix A, and performing classification and identification on the test set of the gait image library by using a K-NN classifier according to the extracted gait features.
And 9, (109) identifying the identity of the underground personnel of the coal mine according to the classification and identification result.
The method is less affected by illumination, can better judge the local information and the category information of the subspace where each high-dimensional data point is located, has high recognition rate and good stability, and can provide reliable information for automatic monitoring of the position of the underground coal mine personnel, automatic identification of the identity of the personnel and the like.
Claims (8)
1. A coal mine underground personnel identity identification method based on gait identification is characterized by comprising the following steps:
A. acquiring gait video sequences of a plurality of operators in a coal mine to form a gait image database;
B. the method for preprocessing the gait video sequence in the gait image database comprises the following steps: firstly, extracting a single frame image from an original gait video image acquired in the step A to perform gray level transformation, then performing median filtering, calculating the median of each pixel point of the gait video sequence after the median filtering frame by frame to be used as a background image of the gait video sequence, extracting a human body target by using a background subtraction method, selecting the gait outline of the human body target by using a rectangular frame with normalized scale, intercepting the gait outline, normalizing the gait outline into 128 multiplied by 64 pixels according to a 1:1 scaling ratio, synthesizing a periodic gait image into a binary image by using a detected gait cycle, converting each image matrix into vector representation to obtain a gait video image vector of a high-dimensional space, and then selecting half of the gait video sequence of each operator as a training set and the other half as a test set;
C. and carrying out dimensionality reduction on the gait video image vectors in the training set, and solving an optimal projection matrix, wherein the solving of the optimal projection matrix is as follows:
(1) the gait video image vector of the high-dimensional space in the training set is subjected to dimension reduction,
(2) constructing a weighted neighborhood graph according to a k-nearest neighbor criterion, setting a neighborhood weight matrix by using a self-adaptive nearest neighbor measurement mode,
(3) defining an objective function according to the neighborhood weight matrix, wherein the objective function is used for ensuring that the type of the same or different nearest neighbor sample points is unchanged after projection, the distance of the same type of points is reduced after projection, the distance of the different types of points is enlarged after projection,
(4) optimizing the objective function, and solving the optimized objective function to obtain an optimal projection matrix;
D. using the optimal projection matrix to project the gait video image data of the test set to a low-dimensional gait image data feature classification space to realize the extraction of the gait feature data;
E. inputting the gait image sequence to be identified into a K-NN classifier, performing classification identification on the test set by using the K-NN classifier according to the extracted gait features, and identifying the identity of the underground coal mine personnel according to the classification identification result.
2. The method for identifying the identity of the underground coal mine personnel according to claim 1, which is characterized in that: in the step A, gait video sequences of operators are collected through a camera, the operators walk at 3 visual angles of the front, the incline and the side relative to the camera respectively, and each operator collects 4 color image sequences at each visual angle.
3. The method for identifying the identity of the underground coal mine personnel according to claim 1, which is characterized in that: in step B, the aspect ratio of the rectangular frame is set to 128: 64.
4. The method for identifying the identity of the underground coal mine personnel according to claim 1, which is characterized in that: let n sample point sets X ═ X of gait video image vectors in high-dimensional space1,x2,…,xnD, dimension D; the sample point set Y of the gait video image vector of the corresponding low-dimensional space is { Y ═ Y1,y2,…,ynD, dimension;
the adaptive nearest neighbor measurement method comprises the following steps: d (x)i,xj)=(xi-xj)TΣ*(xi-xj)
Wherein: w ═ W-1BW-1=W-1/2(W-1/2BW-1/2)W-1/2=W-1/2B*W-1/2
Σ*=W-1/2(W-1/2BW-1/2+δE)W-1/2=W-1/2(B*+δE)W-1/2
Sigma is the covariance matrix, delta is the softening parameter, [ sigma ]*Is a defined covariance matrix, W is the summed intra-class covariance matrix, B is the inter-class covariance matrix, and the element values in W and B are computed from the k nearest neighbors around each point, each point having its corresponding element value in the W matrix and element value in the B matrix.
5. The method for identifying the identity of the underground coal mine personnel as claimed in claim 4, wherein the method comprises the following steps: in step (2), the neighborhood weight matrix H ═ Hij},
In the formula, N (x)i)、N(xj) Are respectively a sample point xi、xjK sets of nearest neighbor points, ciAnd cjIs a sample point xiAnd xjAnd gamma is a control parameter.
7. the method for identifying the identity of the underground coal mine personnel as claimed in claim 6, wherein the method comprises the following steps: the objective function is optimized as:
8. The method for identifying the identity of the underground coal mine personnel as claimed in claim 7, wherein the method comprises the following steps: solving an optimal projection matrix A by decomposing the eigenvalues, and solving k minimum non-zero eigenvalues lambda if k is the dimensionality of the sample point reduction1,λ2,…,λkRespectively corresponding to the feature vector a1,a2,…,akObtaining the optimal projection matrix A ═ a meeting the optimized objective function1,a2,…,ak}。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810936703.3A CN109241870B (en) | 2018-08-16 | 2018-08-16 | Coal mine underground personnel identity identification method based on gait identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810936703.3A CN109241870B (en) | 2018-08-16 | 2018-08-16 | Coal mine underground personnel identity identification method based on gait identification |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109241870A CN109241870A (en) | 2019-01-18 |
CN109241870B true CN109241870B (en) | 2022-02-22 |
Family
ID=65069986
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810936703.3A Expired - Fee Related CN109241870B (en) | 2018-08-16 | 2018-08-16 | Coal mine underground personnel identity identification method based on gait identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109241870B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110399885B (en) * | 2019-07-12 | 2022-05-27 | 武汉科技大学 | Image target classification method based on local geometric perception |
CN110569843B (en) * | 2019-09-09 | 2022-02-08 | 中国矿业大学(北京) | Intelligent detection and identification method for mine target |
CN114359880B (en) * | 2022-03-18 | 2022-05-24 | 北京理工大学前沿技术研究院 | Riding experience enhancement method and device based on intelligent learning model and cloud |
CN115546710A (en) * | 2022-08-09 | 2022-12-30 | 国网湖北省电力有限公司黄龙滩水力发电厂 | Method, device and equipment for locating personnel in hydraulic power plant and readable storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103268500A (en) * | 2013-05-29 | 2013-08-28 | 山东大学 | Gait identifying method with robustness to walking gait changes |
CN103745205A (en) * | 2014-01-22 | 2014-04-23 | 山东大学 | Gait recognition method based on multi-linear mean component analysis |
CN104217211A (en) * | 2014-08-07 | 2014-12-17 | 哈尔滨工程大学 | Multi-visual-angle gait recognition method based on optimal discrimination coupling projection |
CN108121965A (en) * | 2017-12-21 | 2018-06-05 | 深圳大学 | Image-recognizing method based on robust joint sparse feature extraction |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100951890B1 (en) * | 2008-01-25 | 2010-04-12 | 성균관대학교산학협력단 | Method for simultaneous recognition and pose estimation of object using in-situ monitoring |
-
2018
- 2018-08-16 CN CN201810936703.3A patent/CN109241870B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103268500A (en) * | 2013-05-29 | 2013-08-28 | 山东大学 | Gait identifying method with robustness to walking gait changes |
CN103745205A (en) * | 2014-01-22 | 2014-04-23 | 山东大学 | Gait recognition method based on multi-linear mean component analysis |
CN104217211A (en) * | 2014-08-07 | 2014-12-17 | 哈尔滨工程大学 | Multi-visual-angle gait recognition method based on optimal discrimination coupling projection |
CN108121965A (en) * | 2017-12-21 | 2018-06-05 | 深圳大学 | Image-recognizing method based on robust joint sparse feature extraction |
Non-Patent Citations (1)
Title |
---|
基于最大最小判别映射的煤矿井下人员身份鉴别方法;张善文 等;《煤炭学报》;20131031;第1-6页 * |
Also Published As
Publication number | Publication date |
---|---|
CN109241870A (en) | 2019-01-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109241870B (en) | Coal mine underground personnel identity identification method based on gait identification | |
CN108319964B (en) | Fire image recognition method based on mixed features and manifold learning | |
CN106295124B (en) | The method of a variety of image detecting technique comprehensive analysis gene subgraph likelihood probability amounts | |
Kusakunniran et al. | Recognizing gaits across views through correlated motion co-clustering | |
US10853633B2 (en) | Gait recognition system to identify walking subject | |
US7957560B2 (en) | Unusual action detector and abnormal action detecting method | |
US8649594B1 (en) | Active and adaptive intelligent video surveillance system | |
KR101433472B1 (en) | Apparatus, method and computer readable recording medium for detecting, recognizing and tracking an object based on a situation recognition | |
CN105023008A (en) | Visual saliency and multiple characteristics-based pedestrian re-recognition method | |
CN109801305B (en) | SAR image change detection method based on deep capsule network | |
CN106529378B (en) | A kind of the age characteristics model generating method and age estimation method of asian ancestry's face | |
CN112464730B (en) | Pedestrian re-identification method based on domain-independent foreground feature learning | |
CN106096517A (en) | A kind of face identification method based on low-rank matrix Yu eigenface | |
CN113640397B (en) | Remote sensing image change detection method and system based on depth network | |
Yang et al. | Two-stage clustering technique based on the neighboring union histogram for hyperspectral remote sensing images | |
CN115527269B (en) | Intelligent human body posture image recognition method and system | |
CN112990314A (en) | Hyperspectral image anomaly detection method and device based on improved isolated forest algorithm | |
CN106682653A (en) | KNLDA-based RBF neural network face recognition method | |
CN107045624B (en) | Electroencephalogram signal preprocessing and classifying method based on maximum weighted cluster | |
CN106886739A (en) | A kind of video frequency monitoring method based on recognition of face | |
CN109241932B (en) | Thermal infrared human body action identification method based on motion variance map phase characteristics | |
CN101630369A (en) | Pedestrian detection method based on wavelet fractal characteristic | |
CN113591607B (en) | Station intelligent epidemic situation prevention and control system and method | |
CN110674755B (en) | Gait recognition method based on optimum gait flow pattern space | |
CN104615985A (en) | Identification method for person-face similarity |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220222 |