CN107103303A - A kind of pedestrian detection method based on GMM backgrounds difference and union feature - Google Patents
A kind of pedestrian detection method based on GMM backgrounds difference and union feature Download PDFInfo
- Publication number
- CN107103303A CN107103303A CN201710284943.5A CN201710284943A CN107103303A CN 107103303 A CN107103303 A CN 107103303A CN 201710284943 A CN201710284943 A CN 201710284943A CN 107103303 A CN107103303 A CN 107103303A
- Authority
- CN
- China
- Prior art keywords
- image
- pedestrian
- difference
- gmm
- backgrounds
- 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 46
- 238000000034 method Methods 0.000 claims abstract description 30
- 238000012549 training Methods 0.000 claims abstract description 13
- 238000010586 diagram Methods 0.000 claims abstract description 4
- 238000012360 testing method Methods 0.000 claims description 6
- 238000005260 corrosion Methods 0.000 claims description 5
- 230000007797 corrosion Effects 0.000 claims description 5
- 238000012706 support-vector machine Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 230000003628 erosive effect Effects 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 3
- 238000012544 monitoring process Methods 0.000 abstract description 2
- 238000002474 experimental method Methods 0.000 abstract 1
- 238000013461 design Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of pedestrian detection method based on GMM backgrounds difference and union feature, belong to video monitoring and identification technology field.The present invention includes step：First using background modeling of the mixed Gauss model to video image, obtain doing calculus of differences with video present frame after background image, draw the position of sport foreground object, determine the region to be detected of image, the grader for reusing union feature training is treated detection zone and detected, finally obtains the block diagram of pedestrian.The present invention can effectively reduce retrieval number of times of the search domain of grader sliding window with reducing grader, it is determined that the particular location of pedestrian is obtained on the premise of moving target rapidly, method before also having preferable Detection results, experiment to prove relatively for group's pedestrian target has reached more preferable effect.
Description
Technical field
The present invention relates to a kind of pedestrian detection method based on GMM backgrounds difference and union feature, belong to video monitoring with
Identification technology field.
Background technology
In traffic and daily video scene as there are one of most objects in pedestrian, for video automatic identification technology
Research is most important, so pedestrian detection technology turns into study hotspot in recent years, in the field such as intelligent transportation and intelligent security guard hair
Exhibition is rapid.But although the development pedestrian detection technology for passing through the nearly more than ten years achieves certain achievement, but there is presently no
A kind of pedestrian detecting system can be general under any environment and background.
Pedestrian detection method in recent years based on machine learning and characteristic Design has become the main flow in pedestrian detection field
Research direction, the feature of design includes Corner Feature（haar）, LBP（Local binary patterns）, edgelet（Edge feature）, ladder
Spend direction histogram（HOG）Deng, and machine learning algorithm then mainly uses SVMs（SVM）And cascade classifier
（Adaboost）, existing pedestrian detection technology has detection speed slow, and recall rate is not high, the problems such as false drop rate is higher.
The detection of moving object mainly has frame difference method and Background difference in current video sequence, and GMM background subtractions are just
It using mixed Gauss model method is that each pixel in image sets up Gaussian Background model exhibition to be, using going out in video sequence
Existing probability constitutes background image more a little louder, with clear background, the advantages of arithmetic speed can meet real-time.
Gradient orientation histogram（Histogram of Oriented Gradient, HOG）Be one kind in computer vision
With the Feature Descriptor for being used for carrying out object detection in image processing field, he is by calculating the ladder with statistical picture regional area
Degree histogram carrys out constitutive characteristic.HOG description are proposed by French scientist Dalal earliest, but HOG description combine SVM machines
The grader of device learning algorithm training is not high for the correct recall rate of picture, and detection efficiency is relatively low in detection, it is impossible to full
The requirement of real-time in sufficient video.
The content of the invention
The invention provides a kind of pedestrian detection method based on GMM backgrounds difference and union feature, this method is for passing
The grader of the machine learning of system+Feature Descriptor training for video sequence detection speed it is slower the problem of, detect accuracy
It is low, it is proposed that a kind of pedestrian detection method of combination video image object of which movement information, first using mixed Gauss model to video
The background modeling of image, obtains doing calculus of differences with video present frame after background image, draws the position of sport foreground object,
The region to be detected of image is determined, the grader for reusing union feature training is treated detection zone and detected, finally obtains
The block diagram of pedestrian.
Methods described is comprised the following steps that：
Step1, collection video sequence image；
Step2, the sequence image application mixed Gauss model method progress background modeling to being gathered in step Step 1, are carried on the back
Scape image；
Step3, current frame image and background two field picture are made the difference to partite transport calculation, draw differential image region, and to the diff area
Do and corrode, expand, maximum operation, a rectangular area is drawn to the region progress maximization rectangular area of UNICOM, by rectangle
The position in region maps to the position pedestrian window to be detected of present frame after current frame image, i.e. mapping；
Step4, the pedestrian detection grader for using based on HOG and LBP union features and using the training of support vector machines method
The window's position drawn using in step Step3 is detected as ROI on current frame image；
Step5, multiple scale detecting is carried out to pedestrian in present frame window to be detected, detect pedestrian image, and use rectangle frame
Mark.
The method that the background image of the step Step2 is extracted uses the background difference based on mixed Gauss model to calculate
Method.
Differential image region is carried out after corrosion expansive working in the step Step3, sets and mends in maximum operation
Value is repaid, the region after corrosion is subjected to UNICOM, to close differential image regional connectivity, the adjacent difference of offset will be met
Image-region carries out UNICOM.
In described step Step3, the connection condition of adjacent area is that left and right, the distance of neighbouring pixel is respectively less than
Offset 0.03（x+y）, wherein x is the pixel wide of video image, and y is the length in pixels of video image.
In the step Step4, the Cleaning Principle that SVM methods train the model come is the ROI of target image, that is, is felt
Interest region.
In described step Step5, the final presentation in original image of testing result to image.
The beneficial effects of the invention are as follows：
1st, static images are detected with this method is by using movable information to pedestrian relative to previous exclusive use model
Approximate location is judged, reduces the region of window scanning, greatly reduces detection time, improves the efficiency of pedestrian detecting system；
2nd, mixed Gaussian method detection moving target is used alone relative to previous, this method directly determines pedestrian's classification, entered
One step improves the specific aim of moving object detection；
3rd, the extraction that entire area group's pedestrian target can be not added with distinguishing by mixed Gauss model method is used alone relative to previous
Out, this method provides specific position respectively to the single pedestrian target of group pedestrian；
4th, the degree of accuracy of detection is added relative to conventional method, about two percentage points are lifted in the test of common data sets；
5th, single frames test speed of the invention reaches 60ms, and real-time is substantially met in the case where not influenceing Detection accuracy
It is required that.
Brief description of the drawings
Fig. 1 is the flow chart in the present invention；
The design sketch that Fig. 2 obtains for the present invention.
Embodiment
Embodiment 1：As shown in Figure 1-2, a kind of pedestrian detection method based on GMM backgrounds difference and union feature,
First the background modeling of video image is obtained doing difference with video present frame after background image using mixed Gauss model
Computing, draws the position of sport foreground object, determines the region to be detected of image, reuses the grader pair of union feature training
Region to be detected is detected, finally obtains the block diagram of pedestrian.
Methods described is comprised the following steps that：
Step1, collection video sequence image；
Step2, the sequence image application mixed Gauss model method progress background modeling to being gathered in step Step 1, are carried on the back
Scape image；
Step3, current frame image and background two field picture are made the difference to partite transport calculation, draw differential image region, and to the diff area
Do and corrode, expand, maximum operation, a rectangular area is drawn to the region progress maximization rectangular area of UNICOM, by rectangle
The position in region maps to the position pedestrian window to be detected of present frame after current frame image, i.e. mapping；
Step4, the pedestrian detection grader for using based on HOG and LBP union features and using the training of support vector machines method
The window's position drawn using in step Step3 is detected as ROI on current frame image；
Step5, multiple scale detecting is carried out to pedestrian in present frame window to be detected, detect pedestrian image, and use rectangle frame
Mark.
As the preferred scheme of the present invention, the method that the background image of the step Step2 is extracted is used based on mixed
Close the background difference algorithm of Gauss model.
As the preferred scheme of the present invention, differential image region is carried out after corrosion expansive working in the step Step3,
The setting compensation value in maximum operation, UNICOM is carried out by the region after corrosion, will to close differential image regional connectivity
The adjacent differential image region for meeting offset carries out UNICOM.
As the preferred scheme of the present invention, in described step Step3, the connection condition of adjacent area is, left and right, up and down
The distance of adjacent pixel is respectively less than offset 0.03（x+y）, wherein x is the pixel wide of video image, and y is the picture of video image
Plain length.
As the preferred scheme of the present invention, in the step Step4, SVM methods train the Cleaning Principle of the model come
For the ROI of target image, i.e. area-of-interest.
As the preferred scheme of the present invention, in described step Step5, the testing result to image is final in original image
Present.
Embodiment 2：As shown in Figure 1-2, a kind of pedestrian detection method based on GMM backgrounds difference and union feature, such as schemes
Shown in 1-2：Described GMM background modelings difference and the pedestrian detection method of union feature, algorithm are broadly divided into two parts, one
Part be the motion feature for extracting moving target in video as region to be detected, a part uses joint for region to be detected
The grader of features training is detected whether determine moving object is pedestrian.
As shown in Figure 1-2：Described GMM background modelings difference and the pedestrian detection method of union feature, use background subtraction
Sub-model carries out pedestrian movement's feature extraction：Real-time modeling set is carried out to background using GMM methods, background image, Background is obtained
As can interval time be updated, with weaken light change and image in small sample perturbations influence.
As shown in Figure 1-2：Described GMM background modelings difference and the pedestrian detection method of union feature, it is special using joint
The detection that training grader carries out region to be detected is levied, the feature classifiers training sample of training, which is used, includes the positive sample of pedestrian
Do not include the negative sample of pedestrian, sample is trained using the method for SVMs, sorter model is drawn.
As shown in Figure 1-2：Described GMM background modelings difference and the pedestrian detection method of union feature is extracting motion spy
Concretely comprised the following steps during levying, present frame carries out thresholding behaviour with obtaining difference image after background frame difference to difference image
Make, discontinuously the managing to image using morphology closed operation everywhere, obtain incomplete rectangular area, rectangular area is mended
Repay and finally give pedestrian region to be detected.
As shown in Figure 1-2：Described GMM background modelings difference and the pedestrian detection method of union feature realize that effect can
To be detected to the region pedestrian that coincides, the search domain effectively reduced in picture improves detection speed.
The accuracy rate contrast of 1 four kinds of distinct methods of table
The arithmetic speed contrast of 2 four kinds of methods of table
The present invention can effectively reduce the search domain of grader sliding window with reducing the retrieval number of times of grader, it is determined that
The rapid particular location for obtaining pedestrian, also has preferable Detection results for group's pedestrian target on the premise of moving target, tests
Method before proving relatively has reached more preferable effect.
Above in conjunction with accompanying drawing to the present invention embodiment be explained in detail, but the present invention be not limited to it is above-mentioned
Embodiment, can also be before present inventive concept not be departed from the knowledge that those of ordinary skill in the art possess
Put that various changes can be made.
Claims (7)
1. a kind of pedestrian detection method based on GMM backgrounds difference and union feature, it is characterised in that：First use mixed Gaussian mould
Type obtains doing calculus of differences with video present frame after background image, draws sport foreground thing to the background modeling of video image
The position of body, determines the region to be detected of image, and the grader for reusing union feature training is treated detection zone and detected,
Finally obtain the block diagram of pedestrian.
2. the pedestrian detection method according to claim 1 based on GMM backgrounds difference and union feature, it is characterised in that：
Methods described is comprised the following steps that：
Step1, collection video sequence image；
Step2, the sequence image application mixed Gauss model method progress background modeling to being gathered in step Step 1, are carried on the back
Scape image；
Step3, current frame image and background two field picture are made the difference to partite transport calculation, draw differential image region, and to the diff area
Do and corrode, expand, maximum operation, a rectangular area is drawn to the region progress maximization rectangular area of UNICOM, by rectangle
The position in region maps to the position pedestrian window to be detected of present frame after current frame image, i.e. mapping；
Step4, the pedestrian detection grader for using based on HOG and LBP union features and using the training of support vector machines method
The window's position drawn using in step Step3 is detected as ROI on current frame image；
Step5, multiple scale detecting is carried out to pedestrian in present frame window to be detected, detect pedestrian image, and use rectangle frame
Mark.
3. the pedestrian detection method according to claim 1 based on GMM backgrounds difference and union feature, it is characterised in that：
The method that the background image of the step Step2 is extracted uses the background difference algorithm based on mixed Gauss model.
4. the pedestrian detection method according to claim 1 based on GMM backgrounds difference and union feature, it is characterised in that：
Differential image region is carried out after corrosion expansive working in the step Step3, the setting compensation value in maximum operation, by corruption
Region after erosion carries out UNICOM to close differential image regional connectivity.
5. the pedestrian detection method according to claim 1 based on GMM backgrounds difference and union feature, it is characterised in that：
In described step Step3, the connection condition of adjacent area is that left and right, the distance of neighbouring pixel is respectively less than offset
0.03（x+y）, wherein x is the pixel wide of video image, and y is the length in pixels of video image.
6. the pedestrian detection method according to claim 1 based on GMM backgrounds difference and union feature, it is characterised in that：
In the step Step4, the Cleaning Principle that SVM methods train the model come is the ROI of target image, i.e. area-of-interest.
7. the pedestrian detection method according to claim 1 based on GMM backgrounds difference and union feature, it is characterised in that：
In described step Step5, the final presentation in original image of testing result to image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710284943.5A CN107103303A (en) | 2017-04-27 | 2017-04-27 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710284943.5A CN107103303A (en) | 2017-04-27 | 2017-04-27 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107103303A true CN107103303A (en) | 2017-08-29 |
Family
ID=59658182
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710284943.5A Pending CN107103303A (en) | 2017-04-27 | 2017-04-27 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107103303A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108010047A (en) * | 2017-11-23 | 2018-05-08 | 南京理工大学 | A kind of moving target detecting method of combination unanimity of samples and local binary patterns |
| CN108229319A (en) * | 2017-11-29 | 2018-06-29 | 南京大学 | The ship video detecting method merged based on frame difference with convolutional neural networks |
| CN109359549A (en) * | 2018-09-20 | 2019-02-19 | 广西师范大学 | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP |
| CN109389152A (en) * | 2018-08-30 | 2019-02-26 | 广东工业大学 | A kind of fining recognition methods of the vertical pendant object of transmission line of electricity |
| CN109785386A (en) * | 2017-11-14 | 2019-05-21 | 中国电信股份有限公司 | Object identification localization method and device |
| CN111476064A (en) * | 2019-01-23 | 2020-07-31 | 北京奇虎科技有限公司 | Small target detection method and device, computer equipment and storage medium |
| CN112581492A (en) * | 2019-09-27 | 2021-03-30 | 北京京东尚科信息技术有限公司 | Moving target detection method and device |
| CN112883906A (en) * | 2021-03-15 | 2021-06-01 | 珠海安联锐视科技股份有限公司 | Personnel state analysis method based on target detection |
| CN113343820A (en) * | 2021-05-31 | 2021-09-03 | 湖北微特传感物联研究院有限公司 | Pedestrian detection method and device, computer equipment and storage medium |
| CN115147450A (en) * | 2022-09-05 | 2022-10-04 | 中印云端(深圳)科技有限公司 | Moving target detection method and detection device based on motion frame difference image |
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 |
| CN103530638A (en) * | 2013-10-29 | 2014-01-22 | 无锡赛思汇智科技有限公司 | Method for matching pedestrians under multiple cameras |
| CN104268899A (en) * | 2014-09-22 | 2015-01-07 | 河海大学 | Moving object detection method based on frame difference and background difference |
| CN105005773A (en) * | 2015-07-24 | 2015-10-28 | 成都市高博汇科信息科技有限公司 | Pedestrian detection method with integration of time domain information and spatial domain information |
| CN105354548A (en) * | 2015-10-30 | 2016-02-24 | 武汉大学 | Surveillance video pedestrian re-recognition method based on ImageNet retrieval |
| CN106022231A (en) * | 2016-05-11 | 2016-10-12 | 浙江理工大学 | Multi-feature-fusion-based technical method for rapid detection of pedestrian |
-
2017
- 2017-04-27 CN CN201710284943.5A patent/CN107103303A/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 |
| CN103530638A (en) * | 2013-10-29 | 2014-01-22 | 无锡赛思汇智科技有限公司 | Method for matching pedestrians under multiple cameras |
| CN104268899A (en) * | 2014-09-22 | 2015-01-07 | 河海大学 | Moving object detection method based on frame difference and background difference |
| CN105005773A (en) * | 2015-07-24 | 2015-10-28 | 成都市高博汇科信息科技有限公司 | Pedestrian detection method with integration of time domain information and spatial domain information |
| CN105354548A (en) * | 2015-10-30 | 2016-02-24 | 武汉大学 | Surveillance video pedestrian re-recognition method based on ImageNet retrieval |
| CN106022231A (en) * | 2016-05-11 | 2016-10-12 | 浙江理工大学 | Multi-feature-fusion-based technical method for rapid detection of pedestrian |
Non-Patent Citations (4)
| Title |
|---|
| CHENGBIN ZENG 等: "Fast Human Detection Using Mi-SVM and a Cascade of HOG-LBP Features", 《PROCEEDINGS OF 2010 IEEE 17TH INTERNATIONAL CONFERENCE ON IMAGE PROCESSING》 * |
| 商磊 等: "基于密集光流的步态识别", 《大连理工大学学报》 * |
| 徐渊 等: "结合SVM分类器与HOG特征提取的行人检测", 《计算机工程》 * |
| 黄炎 等: "联合梯度直方图和局部二值模式特征的人体检测", 《光学精密工程》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109785386A (en) * | 2017-11-14 | 2019-05-21 | 中国电信股份有限公司 | Object identification localization method and device |
| CN108010047A (en) * | 2017-11-23 | 2018-05-08 | 南京理工大学 | A kind of moving target detecting method of combination unanimity of samples and local binary patterns |
| CN108229319A (en) * | 2017-11-29 | 2018-06-29 | 南京大学 | The ship video detecting method merged based on frame difference with convolutional neural networks |
| CN109389152A (en) * | 2018-08-30 | 2019-02-26 | 广东工业大学 | A kind of fining recognition methods of the vertical pendant object of transmission line of electricity |
| CN109359549A (en) * | 2018-09-20 | 2019-02-19 | 广西师范大学 | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP |
| CN111476064A (en) * | 2019-01-23 | 2020-07-31 | 北京奇虎科技有限公司 | Small target detection method and device, computer equipment and storage medium |
| CN112581492A (en) * | 2019-09-27 | 2021-03-30 | 北京京东尚科信息技术有限公司 | Moving target detection method and device |
| CN112883906A (en) * | 2021-03-15 | 2021-06-01 | 珠海安联锐视科技股份有限公司 | Personnel state analysis method based on target detection |
| CN113343820A (en) * | 2021-05-31 | 2021-09-03 | 湖北微特传感物联研究院有限公司 | Pedestrian detection method and device, computer equipment and storage medium |
| CN115147450A (en) * | 2022-09-05 | 2022-10-04 | 中印云端(深圳)科技有限公司 | Moving target detection method and detection device based on motion frame difference image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107103303A (en) | A kind of pedestrian detection method based on GMM backgrounds difference and union feature | |
| Wei et al. | Multi-vehicle detection algorithm through combining Harr and HOG features | |
| WO2018072233A1 (en) | Method and system for vehicle tag detection and recognition based on selective search algorithm | |
| CN104951784B (en) | A kind of vehicle is unlicensed and license plate shading real-time detection method | |
| Greenhalgh et al. | Recognizing text-based traffic signs | |
| Alvarez et al. | 3D scene priors for road detection | |
| Pan et al. | A robust system to detect and localize texts in natural scene images | |
| CN108304798A (en) | The event video detecting method of order in the street based on deep learning and Movement consistency | |
| CN106384345B (en) | A kind of image detection and flow statistical method based on RCNN | |
| CN104978567A (en) | Vehicle detection method based on scenario classification | |
| CN103310194A (en) | Method for detecting head and shoulders of pedestrian in video based on overhead pixel gradient direction | |
| CN104050684B (en) | A kind of video frequency motion target sorting technique based on on-line training and system | |
| Masita et al. | Pedestrian detection using R-CNN object detector | |
| Deng et al. | Detection and recognition of traffic planar objects using colorized laser scan and perspective distortion rectification | |
| CN103020614A (en) | Human movement identification method based on spatio-temporal interest point detection | |
| Yao et al. | Coupled multivehicle detection and classification with prior objectness measure | |
| Elbamby et al. | Real-time automatic multi-style license plate detection in videos | |
| Hou et al. | A cognitively motivated method for classification of occluded traffic signs | |
| Lin et al. | A traffic sign recognition method based on deep visual feature | |
| CN112347967B (en) | Pedestrian detection method fusing motion information in complex scene | |
| Zhang et al. | Pedestrian detection aided by fusion of binocular information | |
| Lokkondra et al. | DEFUSE: deep fused end-to-end video text detection and recognition | |
| CN107609454A (en) | A kind of method and device of gesture in identification image | |
| Feng et al. | Non-motor vehicle illegal behavior discrimination and license plate detection based on real-time video | |
| Gui et al. | Pedestrian detection based on modified dynamic background using Gaussian mixture models and HOG-SVM detection |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170829 |