CN109359549A - A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP - Google Patents
A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP Download PDFInfo
- Publication number
- CN109359549A CN109359549A CN201811099965.5A CN201811099965A CN109359549A CN 109359549 A CN109359549 A CN 109359549A CN 201811099965 A CN201811099965 A CN 201811099965A CN 109359549 A CN109359549 A CN 109359549A
- Authority
- CN
- China
- Prior art keywords
- feature
- hog
- lbp
- pedestrian
- mixed gaussian
- 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.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 39
- 239000013598 vector Substances 0.000 claims abstract description 30
- 230000004927 fusion Effects 0.000 claims abstract description 18
- 238000000034 method Methods 0.000 claims abstract description 16
- 239000000284 extract Substances 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims abstract description 6
- 238000013100 final test Methods 0.000 claims abstract description 4
- 238000012545 processing Methods 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 238000009472 formulation Methods 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000005259 measurement Methods 0.000 abstract description 2
- 238000012795 verification Methods 0.000 abstract description 2
- 238000012544 monitoring process Methods 0.000 description 11
- 238000000605 extraction Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004381 surface treatment Methods 0.000 description 1
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
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/215—Motion-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20036—Morphological image processing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention proposes a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP, including acquisition pedestrian image, constitutes sample set;The HOG feature and LBP feature for extracting every width figure respectively carry out Fusion Features to the HOG feature and LBP feature and obtain fusion feature vector;SVM classifier is obtained based on fusion feature vector training;Acquire pedestrian's video;Mixed Gaussian background modeling is carried out to pedestrian's video, extracts the motion target area in video;Feature is extracted using HOG_LBP algorithm to motion target area, obtains the feature vector for needing to detect;It needs the feature vector detected to be input in the SVM classifier for described, obtains final testing result.Method of the present invention improves verification and measurement ratio, shortens detection time.
Description
Technical field
The present invention relates to computer and in field of intelligent monitoring more particularly to a kind of based on mixed Gaussian and HOG_LBP
Pedestrian detection method.
Background technique
Pedestrian detection is widely used in computer vision field, such as security protection, intelligent robot, vision monitoring and behavioural analysis
Deng.Due to the diversity of environmental factor complicated and changeable, different shooting angle and human behavior, the accuracy of pedestrian detection
It is to be improved.Figurate rigidity target is different from having, pedestrian can it is different in posture, block and viewing angle variation etc. just
Face rings detection effect.
Pedestrian detecting system is applied to monitoring system, the shortcomings that intelligence may be implemented, make up artificial detection, auxiliary is mentioned
The accuracy rate of height monitoring carries out early warning to abnormal conditions, avoids unnecessary loss.It is fixed that human body is carried out in virtual reality (AR)
Position, i.e., by being modeled to environment captured by camera, in conjunction with the Attitude estimation of pedestrian detection, pedestrian to be accurately located
Position of the pedestrian in world coordinate system can also carry out the suspicious pedestrian in monitoring accurate in conjunction with the application of intelligent monitoring
Positioning, tracking and control, the maturation of pedestrian detection technology can provide superior technique branch for intelligent monitoring and public safety
It holds.Therefore, the research of pedestrian detection algorithm is still an important topic of computer vision field.
Summary of the invention
In view of the foregoing deficiencies of prior art, the purpose of the present invention is to provide one kind to be based on mixed Gaussian and HOG_
The pedestrian detection method of LBP.
In order to achieve the above objects and other related objects, the present invention provides a kind of row based on mixed Gaussian and HOG_LBP
People's detection method, method includes the following steps:
Pedestrian image is acquired, sample set is constituted;
The HOG feature and LBP feature for extracting every width figure respectively carry out Fusion Features to the HOG feature and LBP feature and obtain
To fusion feature vector;
SVM classifier is obtained based on fusion feature vector training;
Acquire pedestrian's video;
Mixed Gaussian background modeling is carried out to pedestrian's video, extracts the motion target area in video;
Feature is extracted using HOG_LBP algorithm to motion target area, obtains the feature vector for needing to detect;
It needs the feature vector detected to be input in the SVM classifier for described, obtains final testing result.
Optionally, standardization processing is carried out to the pedestrian image of acquisition.
Optionally, the HOG feature of each image, including following sub-step are extracted:
Gray level image is converted by the image Jing Guo standardization processing;
Calculate HOG gradient;
Sample image is divided into several fritters, and dimension-reduction treatment is carried out to HOG feature;
Count the gradient orientation histogram of each pixel in each fritter;
Each fritter is normalized;
Institute's directed quantity is linked up into the Vector Groups to be formed, which is HOG feature.
Optionally, the LBP feature for extracting every width figure, including following sub-step:
Sample image is passed through into Sobel edge processing;
Picture is divided into equal-sized grid, each grid is made comparisons with the gray value in grid 3*3 neighborhood, if
8 pixels of surrounding are greater than intermediate threshold, then are 1, are otherwise 0, compare to obtain 8 bits in order;
It is encoded using LBP invariable rotary equivalent formulations.
Optionally, Fusion Features are carried out to the HOG feature and LBP feature and obtain fusion feature vector, method particularly includes:
HOG feature and LBP feature are serially merged, joint histogram is formed.
Optionally, described that mixed Gaussian background modeling is carried out to pedestrian's video, motion target area in video is extracted,
Specifically includes the following steps:
It will test video and carry out mixed Gaussian background modeling;
Constantly automatically update background model;
After mixed Gaussian foreground segmentation, noise is eliminated by median filtering and retains moving target;
By morphologic closed operation, connected domain is formed, then obtains motion target area.
As described above, a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP of the invention, has beneficial below
Effect:
It is of the present invention a kind of based on mixed Gaussian and the improved pedestrian detection method of HOG_LBP, by establishing pedestrian
The database of image obtains the feature vector of image pattern HOG-LBP, is put into svm and obtains a classifier, acquisition testing sample
This, carries out mixed Gaussian background modeling, extracts motion target area, is detected extracting HOG-LBP feature vector.This side
Method improves verification and measurement ratio, shortens detection time.
Detailed description of the invention
In order to which the present invention is further explained, described content, with reference to the accompanying drawing makees a specific embodiment of the invention
Further details of explanation.It should be appreciated that these attached drawings are only used as typical case, and it is not to be taken as to the scope of the present invention
It limits.
Fig. 1 is flow chart described in inventive method;
Fig. 2 is ETH data set figure;
Fig. 3 is pedestrian detection monitoring data collection figure;
Fig. 4 is classics LBP feature extraction schematic diagram.
Specific embodiment
Illustrate embodiments of the present invention below by way of specific specific example, those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for disclosed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints and application, without departing from
Various modifications or alterations are carried out under spirit of the invention.It should be noted that in the absence of conflict, following embodiment and implementation
Feature in example can be combined with each other.
It should be noted that illustrating the basic structure that only the invention is illustrated in a schematic way provided in following embodiment
Think, only shown in schema then with related component in the present invention rather than component count, shape and size when according to actual implementation
Draw, when actual implementation kenel, quantity and the ratio of each component can arbitrarily change for one kind, and its assembly layout kenel
It is likely more complexity.
The present invention is time-consuming for conventional pedestrian's detection method, the low problem of accuracy rate, propose it is a kind of based on mixed Gaussian and
The pedestrian detection method of HOG_LBP characteristics algorithm.This method reduces detection time, improves the accuracy rate of detection.
As shown in Figure 1, the present invention provides one kind based on mixed Gaussian and HOG_LBP pedestrian detection method, the method is such as
Under:
S1. pedestrian image is acquired using high-definition camera, constitutes sample set.It specifically, is 64* at size by image making
128 picture.
S2. HOG feature and LBP Feature Extraction Feature are used each image respectively, it is special to the HOG feature and LBP
Sign carries out Fusion Features and obtains fusion feature vector;
S3. SVM classifier is obtained based on fusion feature vector training;
S4. using monitoring camera in fixed station acquisition pedestrian video, such as parking lot;
S5. mixed Gaussian background modeling is carried out to pedestrian detection video, extracts motion target area in video;
S6. motion target area is extracted into feature using HOG_LBP algorithm, obtains the feature vector for needing to detect;
S7. the feature vector detected will be needed to be input in classifier, obtains final testing result.
Wherein, S1-S3 is training process, and S4-S7 is detection process.
In this present embodiment, described to use each image HOG feature and LBP Feature Extraction Feature respectively, to described
HOG feature and LBP feature carry out Fusion Features and obtain fusion feature vector;Specifically includes the following steps:
S21. it standardizes: converting grayscale image for image;
S22. HOG gradient is calculated:
Gx(x, y)=H (x+1, y)-H (x-1, y)
Gy(x, y)=H (x, y+1)-H (x, y-1)
In formula, Gx(x, y) and Gy(x, y) is horizontal direction gradient and vertical gradient, and H (x, y) is input pixel
Pixel value, θ (x, y) are gradient direction, and G (x, y) is gradient magnitude;
It S23. is that 64*128 divides for 64*128, the fritter of 32*32,16*16, by the dimension-of HOG feature by sample image
3780 dimensions have been reduced to 756 dimensions, to each pixels statistics their gradient orientation histogram in each fritter;
S24. section normalizes, and fritter where the corresponding pixel of vector is normalized,
Wherein, v indicates that the vector after normalization, v indicate the vector before normalization, | |2Indicate 2 rank models of vector
Number, ε indicate the constant of a very little, take 0.01 here;
S25. the vector of upper surface treatment is linked up the Vector Groups to be formed is HOG feature;
S26. sample image cyanines are crossed into Sobel edge processing first.
Picture is divided into equal-sized grid, each grid is made comparisons with the gray value in grid 3*3 neighborhood, if
8 pixels of surrounding are greater than intermediate threshold, then are 1, are otherwise 0, compare to obtain 8 bits in order;Following Fig. 4
It is shown.
S27. it is encoded using LBP invariable rotary equivalent formulations, formula:
S (x) is center decision function is defined as:U(LBPP,R)≤2 are 0-1 conversion in binary-coding
Number is less than or equal to 2 coding,
Wherein, R is radius, and P is to put number, g on neighborhoodcCentered on, gpFor neighborhood point;
S28. HOG feature and LBP feature are serially merged, forms joint histogram.
Specifically, step S28 includes:
Being located at sample space, there are two different space A, B, and corresponding two feature vectors are respectively α ∈ A, β ∈ B,
Then serial fused eigenmatrix is γ=(α, β).
S31. the feature of extraction is put into SVM classifier, obtains a classifier, then training is completed;
The shape information using image is compared with existing technology, it is contemplated that the texture information of human body effectively extracts line
Feature is managed, pedestrian detection speed is improved.
In this present embodiment, described that mixed Gaussian background modeling is carried out to pedestrian detection video, it extracts in video and moves mesh
Mark region;Specifically include following sub-step:
S51. will test video carry out mixed Gaussian background modeling, background pixel respectively by the mixed distribution of K Gauss Lai
Description;xtFor the pixel value of t moment pixel;K is the number of gauss of distribution function, wi,tFor i-th of Gaussian Profile mould of t moment
The weight of type, ui,tFor the mean value of i-th of Gaussian Profile of t moment;∑i,tCovariance matrix;η is Gaussian density function, d xt's
Dimension, P (xt) it is modeling formula.
S52. a most important step is automatically updating for background in background modeling, and model parameter needs not learn to update,
It is matched with the Gaussian Profile met, each new pixel value xtIt is compared as the following formula with current K model, it is new to directly find matching
The distributed model of pixel value, i.e., with the mean bias of the model in 2.5 δi,t-1It is interior.
|xt-ui,t-1|≤2.5δi,t-1
Think that matched mode meets context request if meeting above-mentioned formula condition, taking for i-th of Gauss M of T moment
Value Mi,tIt is denoted as 1, otherwise Mi,t=0.Then the weight of each mode is normalized again, formula is as follows:
wi,t=(1- α) wi,t-1+αMi,t
ui,t=(1- ρ) ui,t-1+ρxt
Wherein, α is weight learning parameter, and ρ is parameter learning rate,For the variance of i-th of gauss of distribution function of t moment;
T is scheduled threshold value, and B is the K Gaussian Profile pixel description after sequence, and B Gaussian Profile is background before selecting
Model:
S53. after mixed Gaussian foreground segmentation, isolated noise point is eliminated by median filtering, can preferably retain fortune
Moving-target marginal information.
S54. by morphologic closed operation, connected domain is formed, then obtains motion target area.
Before obtained moving region image is put into resulting SVM classifier, then it can start to detect pedestrian.
Experimental result:
The advantage of the invention is that using the fusion feature of HOG_LBP, pre-processed on detection video, using mixed
It closes Gauss and carries out foreground segmentation, extract moving target, reduce area of detection, then, with the image zooming-out of diminution HOG_LBP
Feature is extracted gradient information while obtaining pedestrian's local grain information, the accuracy rate of detection improved, when reducing detection
Between, it can accomplish real-time detection with this method.Can widely it apply in intelligent monitoring, such as parking lot, shop etc..
In an experiment, the pedestrian's video for using high definition monitoring camera to acquire is being used as test sample, resolution ratio
For 768*576, totally 600 frame, compares, as shown in table 1 with HOG, LBP, HOG-LBP, CENTRIST.Wherein omission factor is not
The ratio of the pedestrian sample number and pedestrian sample sum that detected;False detection rate is the number that non-pedestrian sample is detected as pedestrian
The ratio of amount and the sum of non-pedestrian sample, detection speed be the frame number of processing per second when detecting sample, as can be seen from Table 1 originally
Method performance is better than other four kinds of methods, and Fig. 2 is experimental result picture.
1 three kinds of algorithm comparisons of table
The above-described embodiments merely illustrate the principles and effects of the present invention, and is not intended to limit the present invention.It is any ripe
The personage for knowing this technology all without departing from the spirit and scope of the present invention, carries out modifications and changes to above-described embodiment.Cause
This, institute is complete without departing from the spirit and technical ideas disclosed in the present invention by those of ordinary skill in the art such as
At all equivalent modifications or change, should be covered by the claims of the present invention.
Claims (6)
1. a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP, which is characterized in that method includes the following steps:
Pedestrian image is acquired, sample set is constituted;
The HOG feature and LBP feature for extracting every width figure respectively carry out Fusion Features to the HOG feature and LBP feature and are melted
Close feature vector;
SVM classifier is obtained based on fusion feature vector training;
Acquire pedestrian's video;
Mixed Gaussian background modeling is carried out to pedestrian's video, extracts the motion target area in video;
Feature is extracted using HOG_LBP algorithm to motion target area, obtains the feature vector for needing to detect;
It needs the feature vector detected to be input in the SVM classifier for described, obtains final testing result.
2. a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP according to claim 1, which is characterized in that
Standardization processing is carried out to the pedestrian image of acquisition.
3. a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP according to claim 2, which is characterized in that
Extract the HOG feature of each image, including following sub-step:
Gray level image is converted by the image Jing Guo standardization processing;
Calculate HOG gradient;
Sample image is divided into several fritters, and dimension-reduction treatment is carried out to HOG feature;
Count the gradient orientation histogram of each pixel in each fritter;
Each fritter is normalized;
Institute's directed quantity is linked up into the Vector Groups to be formed, which is HOG feature.
4. a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP according to claim 2, which is characterized in that
The LBP feature for extracting every width figure, including following sub-step:
Sample image is passed through into Sobel edge processing;
Picture is divided into equal-sized grid, each grid is made comparisons with the gray value in grid 3*3 neighborhood, if surrounding
8 pixels are greater than intermediate threshold, then are 1, are otherwise 0, compare to obtain 8 bits in order;
It is encoded using LBP invariable rotary equivalent formulations.
5. a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP according to claim 1, which is characterized in that
Fusion Features are carried out to the HOG feature and LBP feature and obtain fusion feature vector, method particularly includes: by HOG feature and LBP
Feature is serially merged, and joint histogram is formed.
6. a kind of pedestrian detection method based on mixed Gaussian and HOG_LBP according to claim 1, which is characterized in that
It is described that mixed Gaussian background modeling is carried out to pedestrian's video, motion target area in video is extracted, following step is specifically included
It is rapid:
It will test video and carry out mixed Gaussian background modeling;
Constantly automatically update background model;
After mixed Gaussian foreground segmentation, noise is eliminated by median filtering and retains moving target;
By morphologic closed operation, connected domain is formed, then obtains motion target area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811099965.5A CN109359549A (en) | 2018-09-20 | 2018-09-20 | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811099965.5A CN109359549A (en) | 2018-09-20 | 2018-09-20 | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109359549A true CN109359549A (en) | 2019-02-19 |
Family
ID=65351035
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811099965.5A Pending CN109359549A (en) | 2018-09-20 | 2018-09-20 | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109359549A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110020627A (en) * | 2019-04-10 | 2019-07-16 | 浙江工业大学 | A kind of pedestrian detection method based on depth map and Fusion Features |
CN110120012A (en) * | 2019-05-13 | 2019-08-13 | 广西师范大学 | The video-splicing method that sync key frame based on binocular camera extracts |
CN111798418A (en) * | 2020-06-22 | 2020-10-20 | 电子科技大学 | Wave-absorbing coating speckle defect detection method based on HOG, LBP and GLCM characteristic fusion |
CN111913584A (en) * | 2020-08-19 | 2020-11-10 | 福州大学 | Mouse cursor control method and system based on gesture recognition |
CN113112565A (en) * | 2021-04-08 | 2021-07-13 | 安徽工程大学 | Map construction method based on robot vision closed-loop detection algorithm, storage medium and equipment |
CN113159009A (en) * | 2021-06-25 | 2021-07-23 | 华东交通大学 | Intelligent monitoring and identifying method and system for preventing ticket evasion at station |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102663409A (en) * | 2012-02-28 | 2012-09-12 | 西安电子科技大学 | Pedestrian tracking method based on HOG-LBP |
CN103324955A (en) * | 2013-06-14 | 2013-09-25 | 浙江智尔信息技术有限公司 | Pedestrian detection method based on video processing |
CN104091157A (en) * | 2014-07-09 | 2014-10-08 | 河海大学 | Pedestrian detection method based on feature fusion |
CN105447503A (en) * | 2015-11-05 | 2016-03-30 | 长春工业大学 | Sparse-representation-LBP-and-HOG-integration-based pedestrian detection method |
CN107103303A (en) * | 2017-04-27 | 2017-08-29 | 昆明理工大学 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
CN107491762A (en) * | 2017-08-23 | 2017-12-19 | 珠海安联锐视科技股份有限公司 | A kind of pedestrian detection method |
-
2018
- 2018-09-20 CN CN201811099965.5A patent/CN109359549A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102663409A (en) * | 2012-02-28 | 2012-09-12 | 西安电子科技大学 | Pedestrian tracking method based on HOG-LBP |
CN103324955A (en) * | 2013-06-14 | 2013-09-25 | 浙江智尔信息技术有限公司 | Pedestrian detection method based on video processing |
CN104091157A (en) * | 2014-07-09 | 2014-10-08 | 河海大学 | Pedestrian detection method based on feature fusion |
CN105447503A (en) * | 2015-11-05 | 2016-03-30 | 长春工业大学 | Sparse-representation-LBP-and-HOG-integration-based pedestrian detection method |
CN107103303A (en) * | 2017-04-27 | 2017-08-29 | 昆明理工大学 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
CN107491762A (en) * | 2017-08-23 | 2017-12-19 | 珠海安联锐视科技股份有限公司 | A kind of pedestrian detection method |
Non-Patent Citations (5)
Title |
---|
FANGLIN WANG 等: "Robust and efficient fragments-based tracking usingmean shift", 《ELECTRONICS AND COMMUNICATIONS》 * |
田仙仙 等: "一种改进HOG特征的行人检测算法", 《计算机科学》 * |
疏坤 等: "基于改进的HOG与Sobel_LBP融合的快速行人检测", 《合肥工业大学学报(自然科学版)》 * |
郑锐 等: "基于混合高斯模型与联合特征的行人检测方法", 《传感器与微系统》 * |
龚露鸣 等: "基于混合高斯模型和HOG_SVM的行人检测模型", 《上海大学学报(自然科学版)》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110020627A (en) * | 2019-04-10 | 2019-07-16 | 浙江工业大学 | A kind of pedestrian detection method based on depth map and Fusion Features |
CN110120012A (en) * | 2019-05-13 | 2019-08-13 | 广西师范大学 | The video-splicing method that sync key frame based on binocular camera extracts |
CN110120012B (en) * | 2019-05-13 | 2022-07-08 | 广西师范大学 | Video stitching method for synchronous key frame extraction based on binocular camera |
CN111798418A (en) * | 2020-06-22 | 2020-10-20 | 电子科技大学 | Wave-absorbing coating speckle defect detection method based on HOG, LBP and GLCM characteristic fusion |
CN111913584A (en) * | 2020-08-19 | 2020-11-10 | 福州大学 | Mouse cursor control method and system based on gesture recognition |
CN111913584B (en) * | 2020-08-19 | 2022-04-01 | 福州大学 | Mouse cursor control method and system based on gesture recognition |
CN113112565A (en) * | 2021-04-08 | 2021-07-13 | 安徽工程大学 | Map construction method based on robot vision closed-loop detection algorithm, storage medium and equipment |
CN113159009A (en) * | 2021-06-25 | 2021-07-23 | 华东交通大学 | Intelligent monitoring and identifying method and system for preventing ticket evasion at station |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109359549A (en) | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP | |
CN107767400B (en) | Remote sensing image sequence moving target detection method based on hierarchical significance analysis | |
CN102324025B (en) | Human face detection and tracking method based on Gaussian skin color model and feature analysis | |
CN104268583B (en) | Pedestrian re-recognition method and system based on color area features | |
CN111539273A (en) | Traffic video background modeling method and system | |
CN109685045B (en) | Moving target video tracking method and system | |
CN105046206B (en) | Based on the pedestrian detection method and device for moving prior information in video | |
CN103049751A (en) | Improved weighting region matching high-altitude video pedestrian recognizing method | |
CN106709472A (en) | Video target detecting and tracking method based on optical flow features | |
CN105160310A (en) | 3D (three-dimensional) convolutional neural network based human body behavior recognition method | |
CN104715238A (en) | Pedestrian detection method based on multi-feature fusion | |
CN105528794A (en) | Moving object detection method based on Gaussian mixture model and superpixel segmentation | |
CN110929593A (en) | Real-time significance pedestrian detection method based on detail distinguishing and distinguishing | |
CN105513053B (en) | One kind is used for background modeling method in video analysis | |
CN108416266A (en) | A kind of video behavior method for quickly identifying extracting moving target using light stream | |
CN103310194A (en) | Method for detecting head and shoulders of pedestrian in video based on overhead pixel gradient direction | |
CN105740945A (en) | People counting method based on video analysis | |
CN109919053A (en) | A kind of deep learning vehicle parking detection method based on monitor video | |
CN106529494A (en) | Human face recognition method based on multi-camera model | |
CN106127812A (en) | A kind of passenger flow statistical method of non-gate area, passenger station based on video monitoring | |
CN111539320B (en) | Multi-view gait recognition method and system based on mutual learning network strategy | |
CN107895379A (en) | The innovatory algorithm of foreground extraction in a kind of video monitoring | |
CN110991398A (en) | Gait recognition method and system based on improved gait energy map | |
CN106529441B (en) | Depth motion figure Human bodys' response method based on smeared out boundary fragment | |
CN106570885A (en) | Background modeling method based on brightness and texture fusion threshold value |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190219 |
|
RJ01 | Rejection of invention patent application after publication |