CN101847265A - Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system - Google Patents
Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system Download PDFInfo
- Publication number
- CN101847265A CN101847265A CN 201010152072 CN201010152072A CN101847265A CN 101847265 A CN101847265 A CN 101847265A CN 201010152072 CN201010152072 CN 201010152072 CN 201010152072 A CN201010152072 A CN 201010152072A CN 101847265 A CN101847265 A CN 101847265A
- Authority
- CN
- China
- Prior art keywords
- image
- target
- passenger flow
- moving target
- bus passenger
- 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
Images
Abstract
The invention relates to a new method for extracting moving objects and partitioning multiple objects, in particular to a method for extracting moving passengers and partitioning images applied to bus passenger statistics. The method comprises the following steps of: acquiring images; preprocessing the acquired images; detecting and partitioning moving objects; tracking and calculating the moving objects and the like. A method for extracting the moving objects based on inter-frame difference of edge information can keep the profile of the moving objects, so subsequent processes, such as opening filling, evaluation of the central position of the objects and the like are more correct and more convenient. Multiple moving objects are partitioned by a projection method, so a partition result is correct, and the method is simple and convenient and has high instantaneity.
Description
Technical field
The present invention relates to a kind of new moving target and extract and the multiple goal dividing method, particularly a kind of motion passenger who uses in the bus passenger flow statistics extracts and image partition method.
Background technology
The target detection and the image partition method that are applied to passenger flow statistics have a lot, and the angle of installing according to camera in the passenger flow statistical system can be divided into two classes.One class is for overlooking shooting, can avoid blocking mutually of human body in the image to a certain extent.Another kind of is eyelevel shot, and purpose is the facial information that obtains movement human in image.Camera adopted the situation of overlooking shooting more during moving target extraction of bus passenger flow statistics and target were cut apart.The human detection and the tracking that extract based on head feature proposed in the beach etc., employing is obtained the head degree of depth and see-through feature based on the contouring head feature extraction and the based target parallax that improve the Hough conversion, and they are used to realize having real-time human detection and tracker than the high-accuracy requirement.The precious background frame difference method that proposes based on " minimal gray is poor " of branches of tall trees has effectively been extracted whole pedestrian's moving target.Partly be out of shape minimum characteristic according to the number of people, extract possible number of people curve and carry out template matches, adopt symmetry, head color probability distribution and the movement locus etc. of number of people feature that aim curve is got rid of.Li Zhigang adopts the image resemblance extraction algorithm based on people's face, as unique point, utilizes the method for statistical recognition that the volume of the flow of passengers is added up with the centre coordinate of target area then.S.Harasse etc. have proposed a kind of method of utilizing statistical models to find movement human in video sequence, detect with tracking by skin color modeling, probability shape modeling and Bayes respectively and realize people stream counting.Tarek Yahiaoui etc. has studied a kind of moving object detection of can carrying out on the bigger bus of crowd density, and then finishes the method for bus passenger flow statistics.But above-mentioned target is extracted and the on-the-spot accuracy rate of multiple goal dividing method ubiquity is low, and target is extracted and cut apart shortcoming such as instability, and some algorithm complexity, influences the real-time of system.
Summary of the invention
The present invention existing moving target extract and the basis of target dividing method on, at the site environment of bus passenger flow statistical system, proposes a kind of based on marginal information the extraction of inter-frame difference moving target and based on the multiple goal dividing method of sciagraphy.
Technical scheme of the present invention is: the angle that camera is installed in the bus passenger flow statistical system adopts and overlooks shooting, and the video image of collection is got consecutive frame and is converted to gray level image, through medium filtering, adopts the Canny algorithm to carry out rim detection then.The edge image that obtains is carried out adjacent inter-frame difference, difference image is carried out binary conversion treatment, expansion process is to connect the border lines of the close together that disconnects.The border is filled with scanning algorithm, with the distance that connect the to disconnect border of (15 pixel distances) far away, the edge contour that obtains sealing.Utilize filling algorithm to fill hole in the closed outline then, the mark connected domain is also calculated the area of connected domain, removes little jamming target according to the area size.In bus passenger flow statistical system, the situation because the passenger is blocked, crowded etc. often is connected together by the detected a plurality of moving targets of inter-frame difference, need carry out image segmentation.The present invention adopts to be cut apart based on the multiple goal of sciagraphy, and the image after filling in level, is vertically reached diagonal and carries out projection, obtains pixel count at the x axle, and the projection distribution curve on y axle and the primary and secondary diagonal line is cut apart image according to the peak-to-valley value of curve.Image after cutting apart adopts the center-of-mass coordinate of quaternary tree each cut zone area of algorithm computation and each cut zone, according to the target's center position that calculates, sets up the object chain of moving target, the record object movement locus.The visual field of 320 * 240 pixel images that collect is divided into three zones: the district of getting on the bus, get off district and tracking area.Detect target travel to tracing area and then start the target following chain, be that each object appearing is set up to follow the tracks of chain, and fresh target number status indicator more.To detected new moving target, a newly-built tracking chain, the while is fresh target number status indicator more.When the adjacent three frame averages of the y of target's center coordinate increase gradually, cross counting line, show that the passenger gets off; Otherwise, when the adjacent three frame averages of y coordinate increase for a short time gradually, oppositely cross counting line, show passenger loading, upgrade the counter of getting off of getting on the bus.
Beneficial effect of the present invention is: the method that the inter-frame difference based on marginal information that the present invention proposes extracts moving target can keep the moving target profile, can make follow-up hole fill and ask for target's center position etc. handle more accurate, more convenient.Adopt sciagraphy that multiple mobile object is cut apart, segmentation result is accurate, and method is easy, and real-time is good.
Description of drawings
Fig. 1 is the passenger flow statistical system structured flowchart.
Fig. 2 is a passenger flow statistical system Flame Image Process process flow diagram.
Fig. 3 is moving object detection and extraction algorithm process flow diagram.
Fig. 4 a is first two field picture in adjacent two two field pictures in the bus passenger flow image sequence.
Fig. 4 b is second two field picture in adjacent two two field pictures in the bus passenger flow image sequence.
Fig. 4 c is the movement destination image that adopts the inter-frame difference based on marginal information to obtain.
Fig. 4 d is the movement destination image that adopts traditional inter-frame difference to obtain.
Fig. 5 carries out the movement destination image that filtering obtains to image behind the edge inter-frame difference.
Fig. 6 is the movement destination image that obtains after the border connects.
Fig. 7 is motion target tracking and counting algorithm process flow diagram.
Embodiment:
In the bus passenger flow statistics, use moving target extraction and multiple goal dividing method that the present invention proposes, by analyzing realtime image data, can obtain each website day part number of getting on or off the bus, and this result is fed back to Surveillance center, realize the optimal scheduling of public transport resource.Specific implementation is: the car door place is installed by camera up and down at bus, obtains the image information of passenger getting on/off.The sequential frame image sequence that collects is handled, isolated passenger's target, obtain its details (number, area, position etc.), each passenger who enters the visual field is set up the target following chain, follow the tracks of this passenger,, finally finish automatic counting passengers until walking out the shooting field range.
As shown in Figure 1, bus passenger flow statistical system hardware comprises camera, vehicle-mounted hardware handles platform, several parts such as radio transmitters and Surveillance center's receiver.The CCD camera that is fixed on interior Qianmen of bus and back header obtains image information.The startup of camera is by the door contact interrupter signal controlling, and the passenger getting on/off image sequence that collects is handled through vehicle-mounted hardware handles platform, and result is the number that a certain website is got on or off the bus.Then, pass the ridership information that calculates back Surveillance center by radio transmitters.Surveillance center analyzes the data that receive, and finally realizes the resource rational management.
The camera image acquisition rate is per second 15 frames, and the digital picture size is 320 * 240 pixels.Vehicle-mounted hardware handles platform is mainly finished motion passenger target and is extracted and count tracking, is input as real-time passenger flow view data, is output as the number of getting on or off the bus.TMS320DM6446 processor adopting double-core framework ARM+DSP, wherein arm processor adopts ARM926EJ-S nuclear, and the work dominant frequency is 297MHz, and dsp processor adopts the high-end DSP nuclear C64x+ of TI, the work dominant frequency is 594MHz, can satisfy the requirement that realtime graphic is handled.TMS320DM6446 has internal memory on the sheet, comprise 2 grades of high-speed caches and rich video processing peripheral hardware, and comprise a video/image coprocessor (VICP), and be particularly suitable for development diagram as treatment product, be suitable as the vehicle-mounted hardware platform central processing unit of passenger flow statistical system.
As shown in Figure 2, bus passenger flow statistical system software processes flow process comprises image acquisition, image pre-service, moving object detection and image segmentation, motion target tracking and counting, several parts such as passenger flow output in real time.After getting access to the successive frame sequence image, at first carry out the image pre-service, remove noise, improve picture quality.Moving object detection adopts the frame-to-frame differences point-score based on marginal information, promptly before the difference pretreated image is carried out the Canny rim detection between conducting frame, and the image that will only have marginal information then carries out inter-frame difference, obtains the profile of moving target.Then, the contour curve that obtains being carried out the border connects.Movement destination image after the connection of border is carried out hole fills.Image to multiple mobile object adopts sciagraphy to carry out image segmentation, isolates a plurality of moving targets, calculates clarification of objective value (area, barycenter) then, sets up the target following chain.Judge passenger's the direction of motion and the number of getting on or off the bus, export the volume of the flow of passengers of getting on or off the bus in real time.
As shown in Figure 3, the algorithm flow chart of moving object detection and extraction is input as the successive frame sequence image that collects, and at first carries out the image pre-service, comprises smothing filtering, figure image intensifying.Carrying out pretreated image and need carry out the image type conversion, is 256 grades of gray level images from colored RGB bitmap-converted, to reduce storage and calculation process time.Smothing filtering adopts medium filtering, when removing noise image edge information is not damaged, and can reach satisfied denoising effect.The image enchancing method that adopts is the contrast stretching technology.Rim detection adopts the Canny algorithm, has well kept the marginal information in the image, especially Yun Dong human body contour outline.After the rim detection, gray level image is become bianry image.Handle the result of bus passenger image shown in Fig. 4 a-4d based on the inter-frame difference of marginal information.
Shown in Fig. 4 a-4d, be contrast situation based on marginal information inter-frame difference and traditional inter-frame difference.Employing is based on the inter-frame difference of marginal information, to carrying out the Canny rim detection respectively through the pretreated adjacent two frame bus passenger image graph 4a of image and Fig. 4 b, the edge-detected image that obtains is carried out inter-frame difference, difference result is shown in Fig. 4 c, the moving target that obtains has kept marginal information, is convenient to the follow-up hole that carries out and fills processing and determine the moving target position.And between conventional frame the difference result of method of difference shown in Fig. 4 d, just subtract each other after directly two adjacent two field pictures being done simple pre-service, be easy to the marginal information of moving target is cut, the moving target that lacks marginal information is difficult to accurately obtain its center.
As shown in Figure 5, carry out filtering operation behind the inter-frame difference, adopt median filter method, but the most of noise of filtering.As shown in Figure 6, filtered image is carried out the border connect, can couple together the edge line that disconnects, form the profile of a sealing.The border connects the main two kinds of methods that adopt among the present invention.The morphology expanding processing is used for connecting the point of close together (3-5 pixel), and second method is used on connection level or the vertical direction at interval greater than 5 pixels, less than the point of 15 pixels.The second method concrete steps are as follows: according to from left to right, order traversal bianry image from the top down is zero if the zero pixel value between two adjacent " 1 " pixels of discovery less than 15, is then all composed the value between these two adjacent " 1 " pixels (1).(2) according to from the top down, order from left to right travels through bianry image once more, the same step of pixel value assignment method (1).
As shown in Figure 7, motion target tracking counting process flow diagram at first carries out hole to detected moving target and fills, so that property values such as the area of acquisition moving target, barycenter.Employing is based on the hole filling algorithm of boundary information, at first utilize the border to follow the tracks of the geometric position information that obtains all borders in the image, utilize the geometry site of inner boundary and outer boundary to judge hole location again, at last sweep trace is carried out in the hole zone in the original image and fill.Can there be a plurality of connected components in the image after cutting apart, find out connected components all among the figure, each connected component of difference mark, promptly each connected component all has different mark value.
Multiple goal based on sciagraphy is cut apart, and carries out image segmentation according to the projection distribution characteristics of image on specific direction, and its essence is a kind of statistical method.The employing level, vertically reach three kinds of dividing methods of diagonal.Bianry image after morphology handled is done the projection of horizontal direction, and H represents the result of its horizontal projection, is the one-dimension array that array length equals the bianry image width.
(1)
Wherein M, N represent the width and the height of bianry image respectively, and (x, y) coordinate is (x, the pixel value of y) locating to g in the expression bianry image.
Vertical projection method is that L represents the result of its vertical projection to bianry image projection in vertical direction, is the one-dimension array that array length equals bianry image length.
(2)
To the situation that under horizontal projection and vertical projection situation, all is difficult for cutting apart, can also do projection on principal diagonal and the counter-diagonal direction to bianry image.The projection Z that here defines on the principal diagonal direction is:
Wherein, T was i the pixel on the counter-diagonal and the sum of the one group pixel parallel with the principal diagonal direction.
In like manner, the projection C on the definition counter-diagonal is:
(4)
Wherein, S was i the pixel on the principal diagonal and the sum of the one group pixel parallel with the counter-diagonal direction.
Carrying out multiple goal when cutting apart,, at first image is carried out horizontal projection at the generalized case that bus passenger is got on or off the bus.If there is not tangible peak-to-valley value, again it is carried out the diagonal line projection, if still can't cut apart, carry out vertical projection at last.After obtaining projection, find out projection peak value and valley place coordinate, determine the image segmentation point, be divided into a plurality of moving targets.For each moving target that is partitioned into, calculate its characteristic quantity (area and barycenter) respectively.Mark connected domain and calculate the area of connected domain is fallen some clutter interference targets according to the area size exclusion then.
The visual field of 320 * 240 pixel images that collect is divided into three zones: the district of getting on the bus, get off district and tracking area.The y coordinate of tracking area pixel is 40～200, and the y coordinate is the district of getting on the bus less than 40 the district of getting off that is greater than 200.After obtaining the center-of-mass coordinate of moving target, moving target is followed the tracks of.Track algorithm comprises follows the tracks of startup, data association counting two parts.Before following the tracks of startup, two counters to be set at first, and a status indicator (flag).Status indicator is used for storing the target number of current tracing area, and its initial value is zero.Regulation moving target barycenter y changes in coordinates direction is from big to small for getting on the bus, otherwise for getting off.In case detect target travel to tracing area then start the target following chain, record moving target center-of-mass coordinate value.Being that each object appearing is set up follows the tracks of chain, and fresh target number status indicator more.If existing moving target in the tracing area, currently detect new moving target again, a then newly-built tracking chain, fresh target number status indicator more simultaneously.Data association and segment count receive follow-up consecutive frame image, and the size of target number status indicator flag in target number n and the current tracing area in this two field picture is relatively divided following two kinds of situations:
(1) if | n-flag|=0
Showing does not have fresh target to occur.Can be directly under the single goal situation with new target's center's position renewal in object chain; To the multiple goal situation, then calculate respectively in the present frame target's center position in the target's center position and previous frame apart from d
Ij(i, j are respectively moving target label in present frame and the previous frame).
A) if d
IjGreater than a certain threshold value, then be judged to be miscount, replace calculated value with predicted value, upgrade and follow the tracks of chain;
B) if d
IjIn allowed band, then according to d
IjRespectively with new target's center's position renewal in the object chain of correspondence.
(2) if | n-flag|=1
Showing has fresh target to enter tracing area or target is left tracing area.
A) as if n＞flag, then showing has fresh target to occur, and for the newly-built tracking chain of fresh target, upgrades original object chain;
B) if n＜flag, target is left tracing area, shows that target gets on the bus or get off.Judge that the condition that passenger getting on/off is arranged is that the length of object chain surpasses 15, judge that then the passenger passes through still fare (the y coordinate is 40) down of Top stitch (the y coordinate is 200), upgrades the value of the counter of the corresponding direction of getting on the bus or get off.
Claims (6)
1. a moving target that uses in bus passenger flow statistical system extracts and the multiple goal dividing method, it is characterized in that may further comprise the steps:
(1) video image that camera is photographed is got consecutive frame and is carried out the image pre-service;
(2) image after the pre-service is carried out rim detection, carry out adjacent inter-frame difference then, difference image is carried out processing such as denoising, border connection and hole filling, obtain moving target;
(3) image after will filling adopts sciagraphy to carry out multiple goal to cut apart, calculate the center-of-mass coordinate of each cut zone area and each cut zone, according to the target's center position that calculates, set up the object chain of moving target, the record object movement locus.
2. the moving target that uses in bus passenger flow statistical system according to claim 1 extracts and the multiple goal dividing method, it is characterized in that: detect target travel to tracing area after the described step (3) and then start the target following chain, being that each object appearing is set up follows the tracks of chain, and fresh target number status indicator more.
3. the moving target that uses in bus passenger flow statistical system according to claim 1 and 2 extracts and the multiple goal dividing method, it is characterized in that: camera adopts to overlook to take and obtains video image in the described step (1).
4. the moving target that uses in bus passenger flow statistical system according to claim 1 and 2 extracts and the multiple goal dividing method, and it is characterized in that: described camera image acquisition rate is per second 15 frames, and the digital picture size is 320 * 240 pixels.
5. the moving target that uses in bus passenger flow statistical system according to claim 1 and 2 extracts and the multiple goal dividing method, it is characterized in that: rim detection adopts the Canny algorithm in the described step (2).
6. the moving target that uses in bus passenger flow statistical system according to claim 1 and 2 extracts and the multiple goal dividing method, it is characterized in that: the image after cutting apart in the described step (3) adopts the center-of-mass coordinate of quaternary tree each cut zone area of algorithm computation and each cut zone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010152072 CN101847265A (en) | 2010-04-20 | 2010-04-20 | Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010152072 CN101847265A (en) | 2010-04-20 | 2010-04-20 | Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101847265A true CN101847265A (en) | 2010-09-29 |
Family
ID=42771877
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010152072 Pending CN101847265A (en) | 2010-04-20 | 2010-04-20 | Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101847265A (en) |
Cited By (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102096813A (en) * | 2011-02-10 | 2011-06-15 | 北京邮电大学 | Blob-based pedestrian counting method under vertical visual angle of monocular camera |
| CN102194270A (en) * | 2011-06-02 | 2011-09-21 | 杭州电子科技大学 | Statistical method for pedestrian flow based on heuristic information |
| CN102622798A (en) * | 2012-03-28 | 2012-08-01 | 东南大学 | Passenger flow statistical analysis system |
| CN102855466A (en) * | 2012-04-12 | 2013-01-02 | 无锡慧眼电子科技有限公司 | People counting method based on video image processing |
| CN102855465A (en) * | 2012-04-12 | 2013-01-02 | 无锡慧眼电子科技有限公司 | Tracking method of moving object |
| CN103077403A (en) * | 2012-12-30 | 2013-05-01 | 信帧电子技术（北京）有限公司 | Pedestrian counting method and device |
| CN103428424A (en) * | 2012-05-17 | 2013-12-04 | 佳能株式会社 | Video processing apparatus and method for managing tracking object |
| CN103714698A (en) * | 2013-12-26 | 2014-04-09 | 苏州清研微视电子科技有限公司 | Bus passenger flow statistics system based on distance images |
| CN104933728A (en) * | 2015-07-13 | 2015-09-23 | 天津理工大学 | Mixed motion target detection method |
| CN105100721A (en) * | 2015-08-04 | 2015-11-25 | 中国电子科技集团公司第二十二研究所 | System and transmission method of Beidou intelligent information |
| CN105155971A (en) * | 2015-08-21 | 2015-12-16 | 汪澜 | Automatic door opening system based on gait recognition |
| CN105374048A (en) * | 2014-09-01 | 2016-03-02 | 天津拓视科技有限公司 | Detection method of moving object and system thereof |
| CN105654032A (en) * | 2015-12-15 | 2016-06-08 | 重庆凯泽科技有限公司 | Statistical system of number of people on bus based on face detection and statistical method thereof |
| CN106295523A (en) * | 2016-08-01 | 2017-01-04 | 马平 | A kind of public arena based on SVM Pedestrian flow detection method |
| CN106485265A (en) * | 2016-09-22 | 2017-03-08 | 深圳大学 | A kind of image-recognizing method and device |
| CN106570557A (en) * | 2015-10-13 | 2017-04-19 | 富士通株式会社 | Device and method for counting moving objects |
| CN107240111A (en) * | 2017-06-14 | 2017-10-10 | 郑州天迈科技股份有限公司 | Edge connection segmentation passenger flow statistical method |
| CN107657626A (en) * | 2016-07-25 | 2018-02-02 | 浙江宇视科技有限公司 | The detection method and device of a kind of moving target |
| CN108182230A (en) * | 2017-12-27 | 2018-06-19 | 南京师范大学 | Mobile object convergence pattern method for digging based on bit vector quaternary tree |
| CN108280387A (en) * | 2017-01-05 | 2018-07-13 | 厦门雅迅网络股份有限公司 | Vehicle passenger flow amount statistical method and system |
| CN108334829A (en) * | 2018-01-23 | 2018-07-27 | 浙江大仓信息科技股份有限公司 | A kind of mobilism car passenger number statistical system |
| CN109243150A (en) * | 2018-09-30 | 2019-01-18 | 深圳市金豪泰科技有限公司 | A kind of vehicle early warning method and terminal |
| CN109389007A (en) * | 2017-08-08 | 2019-02-26 | 杭州海康威视数字技术股份有限公司 | Passenger flow volume statistical method, apparatus and system |
| CN109584266A (en) * | 2018-11-15 | 2019-04-05 | 腾讯科技（深圳）有限公司 | A kind of object detection method and device |
| CN109815898A (en) * | 2019-01-24 | 2019-05-28 | 广州市天河区保安服务公司 | Urban district monitoring image acquisition and storage method and system |
| CN109858416A (en) * | 2019-01-21 | 2019-06-07 | 浙江科技学院 | A kind of method of the identification pest and disease damage of insecticidal lamp |
| CN110084197A (en) * | 2019-04-28 | 2019-08-02 | 苏州清研微视电子科技有限公司 | Bus passenger flow volume statistical method and system based on deep learning |
| CN110765940A (en) * | 2019-10-22 | 2020-02-07 | 杭州姿感科技有限公司 | Target object statistical method and device |
| CN111985425A (en) * | 2020-08-27 | 2020-11-24 | 闽江学院 | Image verification device under multi-person scene |
| CN112001334A (en) * | 2020-08-27 | 2020-11-27 | 闽江学院 | Portrait recognition device |
| CN112801018A (en) * | 2021-02-07 | 2021-05-14 | 广州大学 | Cross-scene target automatic identification and tracking method and application |
| CN114119648A (en) * | 2021-11-12 | 2022-03-01 | 史缔纳农业科技(广东)有限公司 | Pig counting method for fixed channel |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1731456A (en) * | 2005-08-04 | 2006-02-08 | 浙江大学 | Bus passenger traffic statistical method based on stereoscopic vision and system therefor |
| CN101321269A (en) * | 2007-06-05 | 2008-12-10 | 同济大学 | Passenger flow volume detection method and system based on computer vision |
| CN101527046A (en) * | 2009-04-28 | 2009-09-09 | 青岛海信数字多媒体技术国家重点实验室有限公司 | Motion detection method, device and system |
-
2010
- 2010-04-20 CN CN 201010152072 patent/CN101847265A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1731456A (en) * | 2005-08-04 | 2006-02-08 | 浙江大学 | Bus passenger traffic statistical method based on stereoscopic vision and system therefor |
| CN101321269A (en) * | 2007-06-05 | 2008-12-10 | 同济大学 | Passenger flow volume detection method and system based on computer vision |
| CN101527046A (en) * | 2009-04-28 | 2009-09-09 | 青岛海信数字多媒体技术国家重点实验室有限公司 | Motion detection method, device and system |
Non-Patent Citations (5)
| Title |
|---|
| 《2009 Second International Conference on Intelligent Networks and Intelligent Systems》 20091231 Yu Ming et al A New Method for Passenger Flow Counting System Based on Surveillance Video 全文 1－6 , 2 * |
| 《ICIAP 2007》 20071231 Massimiliano Bozzoli et al A Statistical Method for People Counting in Crowded Environments 全文 1－6 , 2 * |
| 《ICISP 2008》 20081231 Tarek Yahiaoui et al A People Counting System Based on Dense and Close Stereovision 全文 1－6 , 2 * |
| 《中国图象图形学报》 20091130 朱秋煜 等 一种基于立体视觉的公交车客流计数方法 全文 1－6 第14卷, 第11期 2 * |
| 《光机电研究论坛》 20090228 邵娜 等 视频序列中运动目标检测与跟踪方法研究 第2节第2段1－3行，2.1节2－3行，2.2节第2段1－4行，2.3节第2段，第3节第1段、第2段第6行，以及图1、图2 1－6 , 2 * |
Cited By (47)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102096813A (en) * | 2011-02-10 | 2011-06-15 | 北京邮电大学 | Blob-based pedestrian counting method under vertical visual angle of monocular camera |
| CN102194270A (en) * | 2011-06-02 | 2011-09-21 | 杭州电子科技大学 | Statistical method for pedestrian flow based on heuristic information |
| CN102194270B (en) * | 2011-06-02 | 2012-11-07 | 杭州电子科技大学 | Statistical method for pedestrian flow based on heuristic information |
| CN102622798A (en) * | 2012-03-28 | 2012-08-01 | 东南大学 | Passenger flow statistical analysis system |
| CN102622798B (en) * | 2012-03-28 | 2014-04-16 | 东南大学 | Passenger flow statistical analysis system |
| CN102855465B (en) * | 2012-04-12 | 2016-10-05 | 江苏慧眼数据科技股份有限公司 | A kind of tracking of mobile object |
| CN102855466A (en) * | 2012-04-12 | 2013-01-02 | 无锡慧眼电子科技有限公司 | People counting method based on video image processing |
| CN102855465A (en) * | 2012-04-12 | 2013-01-02 | 无锡慧眼电子科技有限公司 | Tracking method of moving object |
| CN102855466B (en) * | 2012-04-12 | 2016-12-21 | 江苏慧眼数据科技股份有限公司 | A kind of demographic method based on Computer Vision |
| CN103428424B (en) * | 2012-05-17 | 2018-04-10 | 佳能株式会社 | The management method of video processing equipment and tracking object |
| CN103428424A (en) * | 2012-05-17 | 2013-12-04 | 佳能株式会社 | Video processing apparatus and method for managing tracking object |
| CN103077403B (en) * | 2012-12-30 | 2016-12-28 | 信帧电子技术（北京）有限公司 | pedestrian counting method and device |
| CN103077403A (en) * | 2012-12-30 | 2013-05-01 | 信帧电子技术（北京）有限公司 | Pedestrian counting method and device |
| CN103714698B (en) * | 2013-12-26 | 2016-09-21 | 苏州清研微视电子科技有限公司 | Public transit vehicle passenger flow volume statistical system based on range image and method |
| CN103714698A (en) * | 2013-12-26 | 2014-04-09 | 苏州清研微视电子科技有限公司 | Bus passenger flow statistics system based on distance images |
| CN105374048B (en) * | 2014-09-01 | 2019-03-26 | 天津拓视科技有限公司 | The detection method and system of moving target |
| CN105374048A (en) * | 2014-09-01 | 2016-03-02 | 天津拓视科技有限公司 | Detection method of moving object and system thereof |
| CN104933728A (en) * | 2015-07-13 | 2015-09-23 | 天津理工大学 | Mixed motion target detection method |
| CN105100721A (en) * | 2015-08-04 | 2015-11-25 | 中国电子科技集团公司第二十二研究所 | System and transmission method of Beidou intelligent information |
| CN105155971A (en) * | 2015-08-21 | 2015-12-16 | 汪澜 | Automatic door opening system based on gait recognition |
| CN106570557A (en) * | 2015-10-13 | 2017-04-19 | 富士通株式会社 | Device and method for counting moving objects |
| CN105654032A (en) * | 2015-12-15 | 2016-06-08 | 重庆凯泽科技有限公司 | Statistical system of number of people on bus based on face detection and statistical method thereof |
| CN107657626A (en) * | 2016-07-25 | 2018-02-02 | 浙江宇视科技有限公司 | The detection method and device of a kind of moving target |
| CN106295523A (en) * | 2016-08-01 | 2017-01-04 | 马平 | A kind of public arena based on SVM Pedestrian flow detection method |
| CN106485265A (en) * | 2016-09-22 | 2017-03-08 | 深圳大学 | A kind of image-recognizing method and device |
| CN106485265B (en) * | 2016-09-22 | 2020-01-17 | 深圳大学 | Image identification method and device |
| CN108280387A (en) * | 2017-01-05 | 2018-07-13 | 厦门雅迅网络股份有限公司 | Vehicle passenger flow amount statistical method and system |
| CN107240111A (en) * | 2017-06-14 | 2017-10-10 | 郑州天迈科技股份有限公司 | Edge connection segmentation passenger flow statistical method |
| CN109389007A (en) * | 2017-08-08 | 2019-02-26 | 杭州海康威视数字技术股份有限公司 | Passenger flow volume statistical method, apparatus and system |
| CN108182230A (en) * | 2017-12-27 | 2018-06-19 | 南京师范大学 | Mobile object convergence pattern method for digging based on bit vector quaternary tree |
| CN108182230B (en) * | 2017-12-27 | 2020-11-03 | 南京师范大学 | Bit vector quadtree-based mobile object convergence pattern mining method |
| CN108334829B (en) * | 2018-01-23 | 2021-08-06 | 浙江大仓信息科技股份有限公司 | Dynamic passenger car number counting system |
| CN108334829A (en) * | 2018-01-23 | 2018-07-27 | 浙江大仓信息科技股份有限公司 | A kind of mobilism car passenger number statistical system |
| CN109243150A (en) * | 2018-09-30 | 2019-01-18 | 深圳市金豪泰科技有限公司 | A kind of vehicle early warning method and terminal |
| CN109584266A (en) * | 2018-11-15 | 2019-04-05 | 腾讯科技（深圳）有限公司 | A kind of object detection method and device |
| CN109584266B (en) * | 2018-11-15 | 2023-06-09 | 腾讯科技（深圳）有限公司 | Target detection method and device |
| CN109858416A (en) * | 2019-01-21 | 2019-06-07 | 浙江科技学院 | A kind of method of the identification pest and disease damage of insecticidal lamp |
| CN109815898A (en) * | 2019-01-24 | 2019-05-28 | 广州市天河区保安服务公司 | Urban district monitoring image acquisition and storage method and system |
| CN110084197A (en) * | 2019-04-28 | 2019-08-02 | 苏州清研微视电子科技有限公司 | Bus passenger flow volume statistical method and system based on deep learning |
| CN110765940A (en) * | 2019-10-22 | 2020-02-07 | 杭州姿感科技有限公司 | Target object statistical method and device |
| CN112001334A (en) * | 2020-08-27 | 2020-11-27 | 闽江学院 | Portrait recognition device |
| CN111985425A (en) * | 2020-08-27 | 2020-11-24 | 闽江学院 | Image verification device under multi-person scene |
| CN112001334B (en) * | 2020-08-27 | 2024-01-19 | 闽江学院 | Portrait recognition device |
| CN111985425B (en) * | 2020-08-27 | 2024-01-19 | 闽江学院 | Image verification device under multi-person scene |
| CN112801018A (en) * | 2021-02-07 | 2021-05-14 | 广州大学 | Cross-scene target automatic identification and tracking method and application |
| CN112801018B (en) * | 2021-02-07 | 2023-07-07 | 广州大学 | Cross-scene target automatic identification and tracking method and application |
| CN114119648A (en) * | 2021-11-12 | 2022-03-01 | 史缔纳农业科技(广东)有限公司 | Pig counting method for fixed channel |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101847265A (en) | Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system | |
| WO2021208275A1 (en) | Traffic video background modelling method and system | |
| US10445567B2 (en) | Pedestrian head identification method and system | |
| CN102542289B (en) | Pedestrian volume statistical method based on plurality of Gaussian counting models | |
| CN101826228B (en) | Detection method of bus passenger moving objects based on background estimation | |
| US8538082B2 (en) | System and method for detecting and tracking an object of interest in spatio-temporal space | |
| CN104637058B (en) | A kind of volume of the flow of passengers identify statistical methods based on image information | |
| CN103310444B (en) | A kind of method of the monitoring people counting based on overhead camera head | |
| CN110415208B (en) | Self-adaptive target detection method and device, equipment and storage medium thereof | |
| CN103258232B (en) | A kind of public place crowd estimate's method based on dual camera | |
| CN106128121B (en) | Vehicle queue length fast algorithm of detecting based on Local Features Analysis | |
| CN106600625A (en) | Image processing method and device for detecting small-sized living thing | |
| CN103871082A (en) | Method for counting people stream based on security and protection video image | |
| CN101483763A (en) | Digital video processing method oriented to social security monitoring and apparatus thereof | |
| CN104978567B (en) | Vehicle checking method based on scene classification | |
| CN102521565A (en) | Garment identification method and system for low-resolution video | |
| CN104183127A (en) | Traffic surveillance video detection method and device | |
| CN103164858A (en) | Adhered crowd segmenting and tracking methods based on superpixel and graph model | |
| CN103986910A (en) | Method and system for passenger flow statistics based on cameras with intelligent analysis function | |
| CN103593679A (en) | Visual human-hand tracking method based on online machine learning | |
| CN103413330A (en) | Method for reliably generating video abstraction in complex scene | |
| CN105354574A (en) | Vehicle number recognition method and device | |
| CN102289948A (en) | Multi-characteristic fusion multi-vehicle video tracking method under highway scene | |
| CN103473530A (en) | Adaptive action recognition method based on multi-view and multi-mode characteristics | |
| CN109446917A (en) | A kind of vanishing Point Detection Method method based on cascade Hough transform |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20100929 |