CN103150547A - Vehicle tracking method and device - Google Patents
Vehicle tracking method and device Download PDFInfo
- Publication number
- CN103150547A CN103150547A CN2013100215013A CN201310021501A CN103150547A CN 103150547 A CN103150547 A CN 103150547A CN 2013100215013 A CN2013100215013 A CN 2013100215013A CN 201310021501 A CN201310021501 A CN 201310021501A CN 103150547 A CN103150547 A CN 103150547A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- moving region
- current location
- previous moment
- moment position
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a vehicle tracking method and a vehicle tracking device. The vehicle tracking method includes the steps of testing movement of a traffic video image sequence according to a background subtraction method, extracting a movement area including vehicle positions, testing the movement area by adopting a support vector machine (SVM for short) algorithm, obtaining a present position of each vehicle in the movement area, obtaining a position at previous moment of each vehicle, and determining a connecting line of the position at the previous moment and the present position as a movement trail of each vehicle. Due to the vehicle tracking method and the vehicle tracking device, identification process of a traffic video is simplified, construction is easy, and flexibility is high.
Description
Technical field
The present invention relates to the Computer Image Processing field, relate in particular to a kind of wireless vehicle tracking and device.
Background technology
Along with socioeconomic development, the increasing of vehicle, the development of intelligent transportation is also more and more faster, and intelligent transportation system (Intelligent Transportation Sys-tem is referred to as ITS) becomes study hotspot in recent years.Vehicle tracking is based on one of technology of most critical in the intelligent transportation detection system of video technique, is also an important application.The vehicle tracking of video is that the change in time and space of vehicle in video sequence is monitored, and comprises position and movement locus etc., provides foundation for calculating traffic parameter.
But in actual applications, " interlock " problem between many vehicles easily causes the failure of vehicle tracking.Yet for " interlock " problem between many vehicles, correlation technique does not propose effective solution.
Summary of the invention
The present invention proposes a kind of wireless vehicle tracking and device.
In order to achieve the above object, technical scheme of the present invention is achieved in that
According to an aspect of the present invention, provide a kind of wireless vehicle tracking, having comprised: according to the background subtraction method, the traffic video image sequence has been carried out motion detection, extract the moving region that includes vehicle location; Adopt support vector machine (Support Vector Machine is referred to as SVM) algorithm that the moving region is detected, obtain the current location of each vehicle in the moving region; Obtain the previous moment position of each vehicle, and the line between previous moment position and current location is defined as the movement locus of each vehicle.
Preferably, according to the background subtraction method, the traffic video image sequence is carried out motion detection, extract the moving region that includes vehicle location, comprising: the average image that will choose piece image in background or a few width images is image B (x, y) as a setting; Judgement traffic video image sequence I
nPresent frame P (x, y) subtracting background image B (x in (x, y), whether the pixel count of the result images that y) obtains is greater than threshold value, in the situation that judgment result is that and be, determining has the vehicle that is kept in motion in result images, and with result images as moving region Ω
s(x, y).
Preferably, adopt support vector machine (SVM) algorithm that the moving region is detected, obtain the current location of each vehicle in the moving region, comprising: calculate moving region Ω
sThe gradient orientation histogram of all normal windows (HOG) in (x, y); By adopting the good svm classifier device of off-line training to the mode that the feature classification of each HOG judges, determine the current location of each vehicle.
Preferably, after the current location of determining each vehicle, also comprise: the vehicle window of determining to comprise current location.
Preferably, obtain the previous moment position of each vehicle, and the line between previous moment position and current location is defined as the movement locus of each vehicle, comprising: the color histogram that calculates each vehicle window; Default color histogram in each color histogram and predefined template base is compared, obtain comparing result; Identify vehicle in each vehicle window according to comparing result, and extract the previous moment position of the vehicle of sign; With the line of previous moment position and current location as movement locus.
According to another aspect of the present invention, provide a kind of car follower, having comprised: extraction module, be used for according to the background subtraction method, the traffic video image sequence being carried out motion detection, extract the moving region that includes vehicle location; Detection module is used for adopting support vector machine (SVM) algorithm that the moving region is detected, and obtains the current location of each vehicle in the moving region; Acquisition module is used for obtaining the previous moment position of each vehicle, and the line between previous moment position and current location is defined as the movement locus of each vehicle.
Preferably, extraction module comprises: the first setup unit, and the average image that is used for choosing the piece image of background or a few width images is image B (x, y) as a setting; Identifying unit is used for judgement traffic video image sequence I
nPresent frame P (x, y) subtracting background image B (x in (x, y), whether the pixel count of the result images that y) obtains is greater than threshold value, in the situation that judgment result is that and be, determining has the vehicle that is kept in motion in result images, and with result images as moving region Ω
s(x, y).
Preferably, detection module comprises: the first computing unit is used for calculating moving region Ω
sThe gradient orientation histogram of all normal windows (HOG) in (x, y); The first determining unit is used for determining the current location of each vehicle by adopting the good svm classifier device of off-line training to the mode that the feature classification of each HOG judges.
Preferably, detection module also comprises: the second determining unit, and for the vehicle window of determining to comprise current location.
Preferably, acquisition module comprises: the second computing unit, for the color histogram that calculates each vehicle window; The contrast unit is used for the default color histogram of each color histogram and predefined template base is compared, and obtains comparing result; Extraction unit identifies vehicle in each vehicle window according to comparing result, and extracts the previous moment position of the vehicle of sign; The second setup unit is used for line with previous moment position and current location as movement locus.
By the present invention, employing carries out to the traffic video sequence moving region that motion detection is judged the Position Approximate that includes vehicle, vehicle detection is being carried out in the moving region, determine the mode of the movement locus of each vehicle according to the comparing result of the color histogram of the color histogram of the vehicle that extracts and reservation, solved that correlation technique can't solve " interlock " problem between many vehicles and the problem that causes the failure of vehicle tracking, and then reached the identifying of simplifying traffic video, easily construct and the high effect of dirigibility.
Description of drawings
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, the below will do one to the accompanying drawing of required use in embodiment or description of the Prior Art and introduce simply, apparently, accompanying drawing in the following describes is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is the wireless vehicle tracking process flow diagram according to the embodiment of the present invention;
Fig. 2 is wireless vehicle tracking process flow diagram according to the preferred embodiment of the invention;
Fig. 3 is the structured flowchart according to the car follower of the embodiment of the present invention;
Fig. 4 is the structured flowchart of car follower according to the preferred embodiment of the invention.
Embodiment
For the purpose, technical scheme and the advantage that make the embodiment of the present invention clearer, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.
At first, briefly introducing of the wireless vehicle tracking that following examples is provided and car follower:
For specific scene (for example more road conditions of vehicle), first the traffic video sequence of input is carried out motion detection, roughly judge the moving region at the place, position of vehicle; Then, in the moving region that extracts, vehicle is detected; At last, extract and vehicle to have been detected, and this color histogram and the color histogram in the template base of setting up in advance compared, adopt the method for classification to detect each vehicle, each vehicle is followed the tracks of, and then obtained the movement locus of each vehicle.In actual applications, the wireless vehicle tracking that this embodiment provides is identified based on video fully, and construction is easy, and dirigibility is high.
Fig. 1 is the wireless vehicle tracking process flow diagram according to the embodiment of the present invention, and as shown in Figure 1, the method mainly comprises the following steps (step S102-step S106):
Step S102 carries out motion detection according to the background subtraction method to the traffic video image sequence, extracts the moving region that includes vehicle location;
Step S104 adopts support vector machine (SVM) algorithm that the moving region is detected, and obtains the current location of each vehicle in the moving region;
Step S106 obtains the previous moment position of each vehicle, and the line between previous moment position and current location is defined as the movement locus of each vehicle.
In the present embodiment, step S102 can realize by such mode: the average image that will choose piece image in background or a few width images is image B (x, y) as a setting; Judgement traffic video image sequence I
nPresent frame P (x, y) subtracting background image B (x in (x, y), whether the pixel count of the result images that y) obtains is greater than threshold value, in the situation that judgment result is that and be, determining has the vehicle that is kept in motion in result images, and with result images as moving region Ω
s(x, y).
In the present embodiment, step S104 can realize by such mode: calculate moving region Ω
sThe gradient orientation histogram of all normal windows (HOG) in (x, y); By adopting the good svm classifier device of off-line training to the mode that the feature classification of each HOG judges, determine the current location of each vehicle.
Wherein, in step S104, after the current location of determining each vehicle, can also comprise: the vehicle window of determining to comprise current location.
In the present embodiment, step S106 can realize by such mode: the color histogram that calculates each vehicle window; Default color histogram in each color histogram and predefined template base is compared, obtain comparing result; Identify vehicle in each vehicle window according to comparing result, and extract the previous moment position of the vehicle of sign; With the line of previous moment position and current location as movement locus.
For the wireless vehicle tracking that above-described embodiment provides, in actual applications, can adopt in such a way and realize, roughly can adopt following step:
1, adopt the background subtraction method to carry out motion detection to the traffic video image sequence of input, roughly extract the moving region that comprises vehicle location.
2, the moving region to obtaining in step 1 adopts support vector machine method to detect all vehicles, obtains the current location (preferably can also comprise the image size information that vehicle occupies in the moving region) of each vehicle.
3, calculate the color histogram of each vehicle and comparing with the color histogram in the template base of setting up, identify each vehicle, each vehicle current location and previous moment position are linked up and are its movement locus.
Wherein, for step 1, to input video sequence I
n(x, y) present frame is P (x, y), adopt the background subtraction method to detect the moving region: the average image of at first choosing a width in background or a few width images is image B (x as a setting, y), then later sequence image present frame P (x, y) and background image are subtracted each other, carry out the background cancellation.If resulting pixel count greater than a certain threshold value, judges to be monitored that moving object is arranged in scene, thereby obtains moving region Ω
s(x, y).
For step 2, to the moving region Ω that obtains in step 1
s(x, y) employing SVM(support vector machine) method detects all vehicles, obtains the current location (preferably can also comprise the image size information that vehicle occupies in the moving region) of each vehicle:
(1) calculate the gradient orientation histogram (HOG) of the window of all normal sizes in moving region;
(2) adopt the good svm classifier device of off-line training to judge the classification of the HOG feature of each window;
(3) obtain containing the window of vehicle.
For step 3, the color histogram of each the vehicle window that obtains in first calculation procedure 2; Again the color histogram in the color histogram that obtains and the template base of setting up is in advance compared, identified each vehicle according to comparison result; Further, obtain the eve position of the vehicle that has identified, the current location of each vehicle and previous moment position are linked up the movement locus that namely obtains this vehicle.
Below in conjunction with Fig. 2 and preferred embodiment, the wireless vehicle tracking that above-described embodiment provides is further described in more detail.
Fig. 2 is wireless vehicle tracking process flow diagram according to the preferred embodiment of the invention, and as shown in Figure 2, this flow process comprises the following steps (step S202-step S220):
Step S202 reads the traffic video image sequence I of input
n(x, y);
Step S204 is to traffic video image sequence I
nThe present frame P (x, y) of (x, y) adopts the background subtraction method to detect and obtains the moving region, comprising:
At first, the on average image B (x, y) as a setting of a width in background or a few width images will be chosen, then later sequence image present frame P (x, y) and background image are subtracted each other, carry out the background cancellation, obtaining result is the large bianry image T (x, y) such as a width and incoming frame:
Then, seek T (x, y) profile with Findcontours, and then obtain moving region Ω
s(x, y).
Step S206 is for the moving region Ω that obtains
s(x, y) employing SVM(support vector machine) method detects, and detects all vehicles, and obtains current location and/or the size information (this information is for optional) of each vehicle, comprising:
At first, calculate the gradient direction figure of the window of all normal sizes in moving region, be equally divided into nine parts with 360 °, with 1 to 9 nine numeral, mark is carried out in all angles intervals successively.From P (x, y) obtain each normal size coloured image of moving region on, transfer coloured picture to gray-scale map, the x of every and the Grad on the y direction on the calculating gray-scale map, ask for the gradient direction angle of this point, judge which zone is this angle fall in, and then will change the time is labeled as certain number of 1 to 9 accordingly, like this gray-scale map is changed into gradient direction figure, the computing formula of Grad is as follows: Gradient=dy/dx
Dy=g (i, j+1)-g (i, j-1), dx=g (i+1, j)-g (i-1, j), wherein, g (i, j) is the gray-scale value that gray level image is put at (i, j), the Grad of Gradient for calculating.
Then, adopt the good svm classifier device of off-line training to judge the classification of the HOG feature of each window, (wherein positive sample is the image that contains vehicle of normal size to the calculation training sample, negative sample is the image that does not contain vehicle of normal size) gradient direction figure, adopt support vector machine to train the vehicle classification device, according to the classification of the gradient direction figure of the image of each normal size of this sorter judgement moving region;
At last, obtain containing the window of vehicle (showing as vehicle position information).
Step S208 follows the tracks of to each vehicle that (tracing process comprises: step S210-step S220);
Step S210 calculates each vehicle window W obtained above
iThe color histogram of (x, y), this sentences W
1Be example: with W
1The color space of (x, y) is transformed into the hsv space from the rgb space, and h adds up to tone, namely calculates the probability that each color level occurs and obtains color histogram H
1
Step S212 compares color histogram obtained above and the color histogram in the template base of setting up in advance, identifies each vehicle according to comparison result, with color histogram H in the color histogram that obtains and template base
kCoupling, obtain recognition result one by one; Adopt the Pasteur's distance between two color histograms of following formula calculating, the vehicle of matching distance minimum:
Simultaneously, can carry out template base and upgrade: when having new vehicle to enter guarded region, calculate its color histogram and this color histogram is added in template base;
Step S214 links up each vehicle current location and previous moment position and is its movement locus;
Step S216 completes above-mentioned tracing process;
Following steps are (namely in actual applications, carry out namely complete tracing process one time to step S214) alternatively:
Step S218 judges whether present frame is traffic video image sequence I
nThe last frame of (x, y), if so, process ends, if not, execution in step S220.
Fig. 3 is the structured flowchart according to the car follower of the embodiment of the present invention, this device is used for realizing the wireless vehicle tracking that above-described embodiment provides, as shown in Figure 3, this car follower mainly comprises: extraction module 10, detection module 20 and acquisition module 30.Wherein, extraction module 10 is used for according to the background subtraction method, the traffic video image sequence being carried out motion detection, extracts the moving region that includes vehicle location; Detection module 20 is connected to extraction module 10, is used for adopting support vector machine (SVM) algorithm that the moving region is detected, and obtains the current location of each vehicle in the moving region; Acquisition module 30 is connected to detection module 20, is used for obtaining the previous moment position of each vehicle, and the line between previous moment position and current location is defined as the movement locus of each vehicle.
Fig. 4 is the structured flowchart of car follower according to the preferred embodiment of the invention, as shown in Figure 4, in the car follower that the preferred embodiment provides, extraction module 10 comprises: the first setup unit 12, the average image that is used for choosing the piece image of background or a few width images is image B (x, y) as a setting; Identifying unit 14 is connected to the first setup unit 12, is used for judgement traffic video image sequence I
nPresent frame P (x, y) subtracting background image B (x in (x, y), whether the pixel count of the result images that y) obtains is greater than threshold value, in the situation that judgment result is that and be, determining has the vehicle that is kept in motion in result images, and with result images as moving region Ω
s(x, y).
In the car follower that the preferred embodiment provides, detection module 20 comprises: the first computing unit 22 is used for calculating moving region Ω
sThe gradient orientation histogram of all normal windows (HOG) in (x, y); The first determining unit 24 is connected to the first computing unit 22, is used for determining the current location of each vehicle by adopting the good svm classifier device of off-line training to the mode that the feature classification of each HOG judges.
Preferably, detection module 20 can also comprise: the second determining unit 26, be connected to the first determining unit 24, and be used for determining to comprise the vehicle window of current location.
In the car follower that the preferred embodiment provides, acquisition module 30 comprises: the second computing unit 32, for the color histogram that calculates each vehicle window; Contrast unit 34 is connected to the second computing unit 32, is used for the default color histogram of each color histogram and predefined template base is compared, and obtains comparing result; Extraction unit 36 is connected to contrast unit 34, identifies vehicle in each vehicle window according to comparing result, and extracts the previous moment position of the vehicle of sign; The second setup unit 38 is connected to extraction unit 36, is used for line with previous moment position and current location as movement locus.
the wireless vehicle tracking that provides by above-described embodiment and vehicle are according to device, can carry out the moving region that motion detection is judged the Position Approximate that includes vehicle to the traffic video sequence, vehicle detection is being carried out in the moving region, further determine the movement locus of each vehicle according to the comparing result of the color histogram of the color histogram of the vehicle that extracts and reservation, solved that correlation technique can't solve " interlock " problem between many vehicles and the problem that causes the failure of vehicle tracking, and then reached the identifying of simplifying traffic video, easily construct and the high effect of dirigibility.
One of ordinary skill in the art will appreciate that: all or part of step that realizes said method embodiment can be completed by the hardware that programmed instruction is correlated with, aforesaid program can be stored in a computer read/write memory medium, this program is carried out the step that comprises said method embodiment when carrying out; And aforesaid storage medium comprises: the various media that can be program code stored such as ROM, RAM, magnetic disc or CD.
It should be noted that at last: above embodiment only in order to technical scheme of the present invention to be described, is not intended to limit; Although with reference to previous embodiment, the present invention is had been described in detail, those of ordinary skill in the art is to be understood that: it still can be modified to the technical scheme that aforementioned each embodiment puts down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.
Claims (10)
1. a wireless vehicle tracking, is characterized in that, comprising:
According to the background subtraction method, the traffic video image sequence is carried out motion detection, extract the moving region that includes vehicle location;
Adopt the support vector machines algorithm that described moving region is detected, obtain the current location of each vehicle in described moving region;
Obtain the previous moment position of described each vehicle, and the line between described previous moment position and described current location is defined as the movement locus of described each vehicle.
2. method according to claim 1, is characterized in that, according to the background subtraction method, the traffic video image sequence carried out motion detection, extracts the moving region that includes vehicle location, comprising:
With the average image of choosing piece image in background or a few width images image B (x, y) as a setting;
Judge described traffic video image sequence I
n(x, y) the present frame P (x in, y) deduct described background image B (x, whether the pixel count of the result images that y) obtains is greater than threshold value, in the situation that judgment result is that and be, determining has the vehicle that is kept in motion in described result images, and with described result images as described moving region Ω
s(x, y).
3. method according to claim 1 and 2, is characterized in that, adopts the support vector machines algorithm that described moving region is detected, and obtains the current location of each vehicle in described moving region, comprising:
Calculate described moving region Ω
sThe gradient orientation histogram HOG of all normal windows in (x, y);
By adopting the good svm classifier device of off-line training to the mode that the feature classification of each described HOG judges, determine the described current location of described each vehicle.
4. method according to claim 3, is characterized in that, after the described current location of determining described each vehicle, also comprises:
Determine to comprise the vehicle window of described current location.
5. method according to claim 4, is characterized in that, obtains the previous moment position of described each vehicle, and the line between described previous moment position and described current location is defined as the movement locus of described each vehicle, comprising:
Calculate the color histogram of each described vehicle window;
Default color histogram in each described color histogram and predefined template base is compared, obtain comparing result;
Identify vehicle in each described vehicle window according to described comparing result, and extract the described previous moment position of the vehicle of sign;
With the line of described previous moment position and described current location as described movement locus.
6. a car follower, is characterized in that, comprising:
Extraction module is used for according to the background subtraction method, the traffic video image sequence being carried out motion detection, extracts the moving region that includes vehicle location;
Detection module is used for adopting the support vector machines algorithm that described moving region is detected, and obtains the current location of each vehicle in described moving region;
Acquisition module is used for the previous moment position obtain described each vehicle, and the line between described previous moment position and described current location is defined as the movement locus of described each vehicle.
7. device according to claim 6, is characterized in that, described extraction module comprises:
The first setup unit, the average image that is used for choosing the piece image of background or a few width images is image B (x, y) as a setting;
Identifying unit is used for judging described traffic video image sequence I
n(x, y) the present frame P (x in, y) deduct described background image B (x, whether the pixel count of the result images that y) obtains is greater than threshold value, in the situation that judgment result is that and be, determining has the vehicle that is kept in motion in described result images, and with described result images as described moving region Ω
s(x, y).
8. according to claim 6 or 7 described devices, is characterized in that, described detection module comprises:
The first computing unit is used for calculating described moving region Ω
sThe gradient orientation histogram HOG of all normal windows in (x, y);
The first determining unit is used for determining the described current location of described each vehicle by adopting the good svm classifier device of off-line training to the mode that the feature classification of each described HOG judges.
9. device according to claim 8, is characterized in that, described detection module also comprises:
The second determining unit is for the vehicle window of determining to comprise described current location.
10. device according to claim 9, is characterized in that, described acquisition module comprises:
The second computing unit is for the color histogram that calculates each described vehicle window;
The contrast unit is used for the default color histogram of each described color histogram and predefined template base is compared, and obtains comparing result;
Extraction unit identifies vehicle in each described vehicle window according to described comparing result, and extracts the described previous moment position of the vehicle of sign;
The second setup unit is used for line with described previous moment position and described current location as described movement locus.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100215013A CN103150547A (en) | 2013-01-21 | 2013-01-21 | Vehicle tracking method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100215013A CN103150547A (en) | 2013-01-21 | 2013-01-21 | Vehicle tracking method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103150547A true CN103150547A (en) | 2013-06-12 |
Family
ID=48548610
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100215013A Pending CN103150547A (en) | 2013-01-21 | 2013-01-21 | Vehicle tracking method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103150547A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559791A (en) * | 2013-10-31 | 2014-02-05 | 北京联合大学 | Vehicle detection method fusing radar and CCD camera signals |
| CN105354529A (en) * | 2015-08-04 | 2016-02-24 | 北京时代云英科技有限公司 | Vehicle converse running detection method and apparatus |
| CN105488484A (en) * | 2015-12-07 | 2016-04-13 | 北京航空航天大学 | Vehicle track extracting method based on unmanned aerial vehicle image |
| CN106611147A (en) * | 2015-10-15 | 2017-05-03 | 腾讯科技（深圳）有限公司 | Vehicle tracking method and device |
| CN108171989A (en) * | 2018-01-24 | 2018-06-15 | 浙江浩腾电子科技股份有限公司 | It is a kind of to stop capturing system and its application method suitable for the separated of universal ball machine |
| CN109829935A (en) * | 2018-12-29 | 2019-05-31 | 百度在线网络技术（北京）有限公司 | Sequence of scenes tracking processing method, device and the vehicle of vehicle |
| CN110232349A (en) * | 2019-06-10 | 2019-09-13 | 北京迈格威科技有限公司 | Shield lower fingerprint and removes shading method, apparatus, computer equipment and storage medium |
| CN112990002A (en) * | 2021-03-12 | 2021-06-18 | 吉林大学 | Traffic signal lamp identification method and system on downhill road and computer readable medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5761326A (en) * | 1993-12-08 | 1998-06-02 | Minnesota Mining And Manufacturing Company | Method and apparatus for machine vision classification and tracking |
| CN101196996A (en) * | 2007-12-29 | 2008-06-11 | 北京中星微电子有限公司 | Image detection method and device |
| US20110026770A1 (en) * | 2009-07-31 | 2011-02-03 | Jonathan David Brookshire | Person Following Using Histograms of Oriented Gradients |
| CN101976504A (en) * | 2010-10-13 | 2011-02-16 | 北京航空航天大学 | Multi-vehicle video tracking method based on color space information |
| CN102087790A (en) * | 2011-03-07 | 2011-06-08 | 中国科学技术大学 | Method and system for low-altitude ground vehicle detection and motion analysis |
-
2013
- 2013-01-21 CN CN2013100215013A patent/CN103150547A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5761326A (en) * | 1993-12-08 | 1998-06-02 | Minnesota Mining And Manufacturing Company | Method and apparatus for machine vision classification and tracking |
| CN101196996A (en) * | 2007-12-29 | 2008-06-11 | 北京中星微电子有限公司 | Image detection method and device |
| US20110026770A1 (en) * | 2009-07-31 | 2011-02-03 | Jonathan David Brookshire | Person Following Using Histograms of Oriented Gradients |
| CN101976504A (en) * | 2010-10-13 | 2011-02-16 | 北京航空航天大学 | Multi-vehicle video tracking method based on color space information |
| CN102087790A (en) * | 2011-03-07 | 2011-06-08 | 中国科学技术大学 | Method and system for low-altitude ground vehicle detection and motion analysis |
Non-Patent Citations (2)
| Title |
|---|
| 赵建云: "基于视频的交通路口车辆检测技术研究", 《中国优秀硕士学位论文全文数据库》 * |
| 陈功等: "鲁棒的实时多车辆检测与跟踪系统设计", 《信号处理》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559791B (en) * | 2013-10-31 | 2015-11-18 | 北京联合大学 | A kind of vehicle checking method merging radar and ccd video camera signal |
| CN103559791A (en) * | 2013-10-31 | 2014-02-05 | 北京联合大学 | Vehicle detection method fusing radar and CCD camera signals |
| CN105354529A (en) * | 2015-08-04 | 2016-02-24 | 北京时代云英科技有限公司 | Vehicle converse running detection method and apparatus |
| CN106611147B (en) * | 2015-10-15 | 2018-10-16 | 腾讯科技（深圳）有限公司 | Car tracing method and apparatus |
| CN106611147A (en) * | 2015-10-15 | 2017-05-03 | 腾讯科技（深圳）有限公司 | Vehicle tracking method and device |
| CN105488484A (en) * | 2015-12-07 | 2016-04-13 | 北京航空航天大学 | Vehicle track extracting method based on unmanned aerial vehicle image |
| CN105488484B (en) * | 2015-12-07 | 2018-11-27 | 北京航空航天大学 | A kind of track of vehicle extracting method based on unmanned plane image |
| CN108171989A (en) * | 2018-01-24 | 2018-06-15 | 浙江浩腾电子科技股份有限公司 | It is a kind of to stop capturing system and its application method suitable for the separated of universal ball machine |
| CN109829935A (en) * | 2018-12-29 | 2019-05-31 | 百度在线网络技术（北京）有限公司 | Sequence of scenes tracking processing method, device and the vehicle of vehicle |
| US11062150B2 (en) | 2018-12-29 | 2021-07-13 | Baidu Online Network Technology (Beijing) Co., Ltd. | Processing method and apparatus for vehicle scene sequence tracking, and vehicle |
| CN110232349A (en) * | 2019-06-10 | 2019-09-13 | 北京迈格威科技有限公司 | Shield lower fingerprint and removes shading method, apparatus, computer equipment and storage medium |
| CN110232349B (en) * | 2019-06-10 | 2020-07-03 | 北京迈格威科技有限公司 | Under-screen fingerprint shading method, device, computer equipment and storage medium |
| CN112990002A (en) * | 2021-03-12 | 2021-06-18 | 吉林大学 | Traffic signal lamp identification method and system on downhill road and computer readable medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103150547A (en) | Vehicle tracking method and device | |
| Wu et al. | Lane-mark extraction for automobiles under complex conditions | |
| CN105930791B (en) | Recognition method of road traffic signs based on multi-camera fusion based on DS evidence theory | |
| CN106203398B (en) | A kind of method, apparatus and equipment detecting lane boundary | |
| US8902053B2 (en) | Method and system for lane departure warning | |
| US20190122059A1 (en) | Signal light detection | |
| CN112307989B (en) | Road surface object identification method, device, computer equipment and storage medium | |
| CN104616502B (en) | Car license recognition and alignment system based on combination type bus or train route video network | |
| US20070041614A1 (en) | Road marking recognition apparatus and method | |
| CN102915433B (en) | Character combination-based license plate positioning and identifying method | |
| US9152856B2 (en) | Pedestrian detection system and method | |
| CN111325769A (en) | Target object detection method and device | |
| CN110718061A (en) | Traffic intersection traffic flow statistics method, device, storage medium and electronic device | |
| CN111191611A (en) | A traffic sign recognition method based on deep learning | |
| US20170017836A1 (en) | Character segmenting apparatus, character recognition apparatus, and character segmenting method | |
| JP2019220129A (en) | System and method for recognizing traffic sign | |
| Danescu et al. | Detection and classification of painted road objects for intersection assistance applications | |
| EP2813973A1 (en) | Method and system for processing video image | |
| CN103106409A (en) | Composite character extraction method aiming at head shoulder detection | |
| CN111967396A (en) | Processing method, device and equipment for obstacle detection and storage medium | |
| JP5423764B2 (en) | Moving object detection apparatus, computer program, and moving object detection method | |
| CN107133588A (en) | Vehicle identification method based on vehicle window feature extraction | |
| CN106780727A (en) | A kind of headstock detection model method for reconstructing and device | |
| WO2010019024A2 (en) | Method and system for tracking and tagging objects | |
| Huu et al. | Proposing Lane and Obstacle Detection Algorithm Using YOLO to Control Self‐Driving Cars on Advanced Networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130612 |