CN102496282B - Traffic intersection signal light state identification method based on RGB color transformation - Google Patents
Traffic intersection signal light state identification method based on RGB color transformation Download PDFInfo
- Publication number
- CN102496282B CN102496282B CN201110421718.4A CN201110421718A CN102496282B CN 102496282 B CN102496282 B CN 102496282B CN 201110421718 A CN201110421718 A CN 201110421718A CN 102496282 B CN102496282 B CN 102496282B
- Authority
- CN
- China
- Prior art keywords
- image
- lamp
- region
- signal lamp
- pixel
- 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.)
- Active
Links
Images
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a traffic intersection signal light state identification method based on video image processing. The method comprises the following steps of: acquiring a video image in real time via a video device, manually drawing the position of a signal light group on the video image, segmenting the region of the signal lights, and making the region into a mask; after acquiring each frame of image, smoothing the image in the mask region, removing noise influence, and analyzing the signal light image in the mask. A color transformation method is designed for graying color images, namely, three components in the color image are extracted, the min(r-g, r-b) operation is executed on a red or yellow signal light region, and for a green signal light region, the min(g-r, g-b) operation is used. The pixels of the traffic signal lights are separated from the image through a threshold method, the number of pixels separated from 25 frames of image in one second of three signal light regions is respectively accumulated, and the maximum one of the three accumulated values is taken as the basis for judging that the light is on.
Description
Technical field
The present invention designs field of traffic, particularly relates to a kind of traffic intersection signal lamp condition judgement method of processing based on video image, for judging belisha beacon real-time status.
Background technology
For whether effective monitoring driver observes traffic rules and regulations, in existing traffic surveillance and control system, as made a dash across the red light automatic monitoring system, usually need to use red signal.Current way is from teleseme, the red signal of corresponding lamp group to be extracted, and after corresponding red signal transponder, then delivers to crossing and makes a dash across the red light in automatic monitoring system, when current demand signal being detected and being red light, just judges running red light for vehicle state.This way has three deficiencies: the first, and engineering construction trouble, cost improves, and some teleseme red signal is 220V voltage, dangerous; The second, this structure relies on red signal transponder, and multichannel red signal may bring fault etc., and the system that causes making a dash across the red light lacks the key condition of setting out, and the 3rd, signal trouble may bring mistake to clap and leak and clap.
Also some system is used the method that video image is processed to detect traffic intersection signal lamp state at present, by the colouring information of detection signal lamp image red, yellow, and green, and judgement signal lamp state.In order to reach more stable recognition effect, conventionally image can be converted to the color spaces such as HSV, color component is wherein analyzed.Or the color center such as image setting red, yellow, and green to signal lamp region, the method for cutting apart with image obtains the judgement of signal lamp state.This way is usually easily subject to outdoor weather situation impact, outdoor various changeable state of weather cause take color segmentation as main method stability not strong, weather effect usually causes cutting apart failure, the motion of vehicle simultaneously and vibrations also can make image blurring.
The method that adopts video image to process judges the state that teleseme is current, difficulty is: video record equipment works in outdoor, will meet with various situations, variation as strong and weak in illumination, rotation round the clock, rain, snow etc., these all will produce considerable influence to video image.This just requires the corresponding traffic signal light condition discrimination model can self-adaptation, has enough robustnesss, in various environment, can normally work.
In addition, current trend is in web camera, to realize the monitoring function of various crossings vehicle-state, as the detection all-in-one that makes a dash across the red light, and signal lamp identification judgement is the sub-fraction of function just, this just requires signal lamp recognizer to want simple and efficient, can not take too many resource.
Summary of the invention
For solving the problems of the technologies described above, the traffic intersection signal lamp state identification method of a kind of RGB colour switching of the present invention is to adopt following technical scheme to realize.
First, manual signalization lamp lamp group position on the video image gathering, a lamp group comprises three signal lamps of red, yellow, and green, conventionally be arranged as from top to bottom or from left to right, be partitioned into signal lamp region and be made into mask, every collection one two field picture, is partitioned into masks area image, follow-uply take a lamp group and describe as processing unit, a plurality of lamp groups adopt identical method to process.
Secondly, the masks area image of cutting apart is done to Gauss's template smoothing processing of 3 * 3, removed the signal lamp image in noise effect post analysis mask.
Again, design a kind of colour switching method by coloured image gray processing, in the time of speed up processing, utilize the green three kinds of signal lamp color features of reddish yellow, reach stable segmentation effect; Be three component rgb values extracting each pixel in signal lamp image, for redness or amber lamp region, take min (r-g, r-b) computing, for greensignal light region, take min (g-r, g-b) computing, reaches the effect of coloured image gray processing.
Again, the gray level image obtaining is cut apart, obtaining may be the foreground pixel point of bright light; To single pass red eye image, the pixel that is greater than threshold value 1 in image, be red light bright light pixel; To single pass amber lamp image, the pixel that is greater than threshold value 1 in image, be amber light bright light pixel; To single pass greensignal light image, the pixel that is greater than threshold value 1 in image, be green light bright light pixel.
Finally, three pixel number sums that signal lamp region is cut apart in 25 two field pictures for 1 second of accumulative total, using the peaked judgement lighting as this lamp in three.
Figure 1 shows that the process flow diagram of the invention process routine.
Embodiment
In order more to understand technology contents of the present invention, especially exemplified by preferred embodiment, and coordinate appended graphic being described as follows.
With reference to figure 1, Figure 1 shows that the process flow diagram of a preferred embodiment of the present invention.From Fig. 1, we can find out that the traffic intersection signal lamp state identification method of a kind of RGB colour switching of the present invention comprises the following steps.
Step 100: manual signalization lamp lamp group position on the video image gathering, a lamp group comprises three signal lamps of red, yellow, and green, conventionally be arranged as from top to bottom or from left to right, be partitioned into signal lamp region and be made into mask, every collection one two field picture, be partitioned into masks area image, follow-uply take a lamp group and describe as processing unit, a plurality of lamp groups adopt identical method to process; Common image transmitting process can be brought various noise effects, and masks area image is done to Gaussian smoothing, removes noise effect.Gaussian filter is according to the shape of Gaussian function, to select the linear smoothing wave filter of weights.Gaussian filter is very resultful to removing the noise of Normal Distribution.Concerning image, the conventional discrete Gaussian function of two-dimentional zero-mean is made smoothing filter, the masks area image of cutting apart is done to Gauss's template smoothing processing of 3 * 3, removes the signal lamp image in noise effect post analysis mask.
Step 200: method for designing is red colored lamp area grayscale; What step 100 obtained is RGB triple channel coloured image, and system is carried out pre-service by the method for coloured image gray processing to the traffic lights image gathering; Extract three components in coloured image, for the every bit in red eye region, take min (r-g, r-b) computing, get the little value in the difference of r-g and r-b, if be less than 0, be made as 0 value, the gray-scale value using the value obtaining as newly-generated image; After whole processes pixel are complete, just complete coloured image gray processing.
Step 300: method for designing is amber light area grayscale; Extract three components in coloured image, for the every bit in amber lamp region, take min (r-g, r-b) computing, get the little value in the difference of r-g and r-b, if be less than 0, be made as 0 value, the gray-scale value using the value obtaining as newly-generated image.
Step 400: method for designing is green light area grayscale; Extract three components in coloured image, for the every bit in greensignal light region, take min (g-r, g-b) computing, get the little value in the difference of g-r and g-b, if be less than 0, be made as 0 value, the gray-scale value using the value obtaining as newly-generated image.
The preprocess method that two Color Channels subtract each other is (255,0 to the desirable value of color of red eye, 0), the desirable value of color of amber lamp is (255,255,0), greensignal light ideal value is (0,0,255), be corresponding min (r-g, r-b) computing and computing, the color contrast value that the various different colours signal lamps of outstanding behaviours are different, robustness is higher.
This pre-service to the traffic lights image gathering can improve operation efficiency to a great extent, reduces the subsequent treatment time.Because the various impacts in the external world are identical to the effect of each Color Channel of signal lamp, two Color Channels subtract each other simultaneously, can maximum possible offset the impact of extraneous various interference on image.
Step 500: with the gray-scale map of Threshold segmentation red channel, statistics foreground pixel number.
Analyze each area grayscale image that pre-service obtains, in minute spirogram that comprises traffic signal light condition, the brightness of traffic lights position is apparently higher than other regions; This is can be so that the pixel region thresholding of the non-bright light in image is tending towards the little value of gray level more after minimizing operation owing to subtracting each other with the most significant component of this signal lamp and other two components, the pixel region thresholding of bright light keeps larger, thereby makes signal lamp more easily separated.
To red light region, setting threshold is Th1=40, and the point that is greater than threshold value is the pixel of red light region bright light, statistical pixel point number.
Step 600: to amber light region, setting threshold is Th2=40, the point that is greater than threshold value is the pixel of amber light region bright light, statistical pixel point number.
Step 700: to green light region, setting threshold is Th3=40, the point that is greater than threshold value is the pixel of green light region bright light, statistical pixel point number.
Step 800: compare three kinds of signal lamp foreground pixel numbers of 1 second accumulative total, judgement signal lamp state.
To in every two field picture for the signal lamp range statistics bright light pixel number in region, to video frame rate, be the camera of 25 frame/seconds, take and signal lamp state is judged as the cycle for 1 second; The pixel number that three signal lamp regions of accumulative total are cut apart in 25 two field pictures for 1 second is respectively SumPix1, SumPix2, SumPix3, using the peaked judgement of the bright light as this group signal lamp in three, the method can be effectively, the stable judgement that obtains signal lamp state, avoids identifying the identification error that occurs turn-off in the process of stroboscopic signal lamp bright light and cause; Lighting lamp state with same method judgement to other group signal lamps.
Claims (4)
1. a traffic intersection signal lamp state identification method for RGB colour switching, gathers video image in real time by video equipment, it is characterized in that:
1) manual signalization lamp lamp group position on the video image gathering, is partitioned into signal lamp region and is made into mask, and every collection one two field picture, is partitioned into masks area image;
2) the masks area image of cutting apart is done to smoothing processing, remove the signal lamp image in noise effect post analysis mask;
3) design a kind of colour switching method by coloured image gray processing, extract three components in coloured image, for redness or amber lamp region, take min (r-g, r-b) computing, for greensignal light region, take min (g-r, g-b) computing;
4) with threshold method, process traffic lights pixel is split from image;
5) three pixel numbers that signal lamp region is cut apart in 25 two field pictures for 1 second of accumulative total, using the peaked judgement lighting as this lamp in three;
Wherein:
Manual signalization lamp lamp group position on the video image gathering, a lamp group comprises three signal lamps of red, yellow, and green, is conventionally arranged as from top to bottom or from left to right;
Be partitioned into signal lamp region, be every kind of signal lamp zoning in lamp group, and be made into mask;
Every collection one two field picture, is partitioned into mask lamp group area image, follow-uply take a lamp group as processing unit, and a plurality of lamp groups adopt identical method to process.
2. the traffic intersection signal lamp state identification method of a kind of RGB colour switching according to claim 1, is characterized in that:
The lamp group area image of cutting apart is done to Gauss's template smoothing processing of 3 * 3, removed the signal lamp image in noise effect post analysis mask.
3. the traffic intersection signal lamp state identification method of a kind of RGB colour switching according to claim 1, is characterized in that:
1) design a kind of colour switching method by coloured image gray processing, in the time of speed up processing, utilize the green three kinds of signal lamp color features of reddish yellow, reach stable segmentation effect;
2) extract three component r, g, the b value of each pixel in signal lamp image, for redness or amber lamp region, take min (r-g, r-b) computing, for greensignal light region, take min (g-r, g-b) computing.
4. the traffic intersection signal lamp state identification method of a kind of RGB colour switching according to claim 3, is characterized in that:
1) by the red eye image obtaining after claim 3 processing, be single channel image, the pixel that is greater than threshold value 1 in judgement image is red light bright light pixel;
2) by the amber lamp image obtaining after claim 3 processing, be single channel image, the pixel that is greater than threshold value 2 in judgement image is amber light bright light pixel;
3) by the greensignal light image obtaining after claim 3 processing, be single channel image, the pixel that is greater than threshold value 3 in judgement image is green light bright light pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110421718.4A CN102496282B (en) | 2011-12-16 | 2011-12-16 | Traffic intersection signal light state identification method based on RGB color transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110421718.4A CN102496282B (en) | 2011-12-16 | 2011-12-16 | Traffic intersection signal light state identification method based on RGB color transformation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102496282A CN102496282A (en) | 2012-06-13 |
| CN102496282B true CN102496282B (en) | 2014-04-16 |
Family
ID=46188101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110421718.4A Active CN102496282B (en) | 2011-12-16 | 2011-12-16 | Traffic intersection signal light state identification method based on RGB color transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102496282B (en) |
Families Citing this family (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104769652B (en) * | 2012-11-20 | 2017-10-03 | 哈曼国际工业有限公司 | Method and system for detecting traffic lights |
| CN103177256B (en) * | 2013-04-02 | 2016-12-28 | 上海理工大学 | Display state of traffic signal lamp recognition methods |
| CN103324943A (en) * | 2013-06-18 | 2013-09-25 | 中国人民解放军第二炮兵工程大学 | Identification method of complex device panel image multi-sub zone state |
| EP3036730B1 (en) * | 2013-08-20 | 2019-12-18 | Harman International Industries, Incorporated | Traffic light detection |
| CN103632559B (en) * | 2013-12-05 | 2016-03-16 | 武汉烽火众智数字技术有限责任公司 | Based on the lights state detection method of video analysis |
| CN104021378B (en) * | 2014-06-07 | 2017-06-30 | 北京联合大学 | Traffic lights real-time identification method based on space time correlation Yu priori |
| CN105812674B (en) * | 2014-12-29 | 2019-05-21 | 浙江大华技术股份有限公司 | A kind of color correct method, monitoring method and the device of signal lamp |
| MX360623B (en) * | 2015-04-08 | 2018-11-09 | Nissan Motor | Traffic light detection device and traffic light detection method. |
| CN105303860B (en) * | 2015-10-22 | 2017-06-16 | 四川膨旭科技有限公司 | Traffic lights are identified system in vehicle travel process |
| CN105791710B (en) * | 2016-04-11 | 2019-10-29 | 北京英泰智科技股份有限公司 | A kind of signal lamp image enhancement processing method |
| CN107038420A (en) * | 2017-04-14 | 2017-08-11 | 北京航空航天大学 | A kind of traffic lights recognizer based on convolutional network |
| CN108961357B (en) * | 2017-05-17 | 2023-07-21 | 浙江宇视科技有限公司 | Method and device for strengthening over-explosion image of traffic signal lamp |
| CN107507142B (en) * | 2017-08-11 | 2021-02-09 | 江苏航天大为科技股份有限公司 | Signal lamp image processing method |
| CN107749055A (en) * | 2017-09-20 | 2018-03-02 | 广东方纬科技有限公司 | A kind of fault detection method, system and the device of LED traffic guidances screen |
| CN108257401B (en) * | 2018-01-31 | 2021-06-25 | 张转 | Device and method for preventing traffic light of automobile from being unclear due to direct sunlight |
| CN108391602B (en) * | 2018-04-25 | 2023-06-06 | 中国农业科学院农业信息研究所 | Chick gender identification system and chick gender identification method |
| CN111402610B (en) * | 2020-03-23 | 2022-05-10 | 东软睿驰汽车技术（沈阳）有限公司 | Method, device, equipment and storage medium for identifying lighting state of traffic light |
| CN112180285B (en) * | 2020-09-23 | 2024-05-31 | 阿波罗智联(北京)科技有限公司 | Method and device for identifying traffic signal lamp faults, navigation system and road side equipment |
| CN111950536B (en) * | 2020-09-23 | 2024-08-02 | 阿波罗智联(北京)科技有限公司 | Signal lamp image processing method and device, computer system and road side equipment |
| CN114038204A (en) * | 2020-10-21 | 2022-02-11 | 南京光准科技有限公司 | Device for realizing early warning of red light running risk by sensing traffic light state through laser |
| CN113345036B (en) * | 2021-05-24 | 2022-09-27 | 广西电网有限责任公司电力科学研究院 | HSV (hue, saturation, value) feature transformation based indicator lamp state identification method |
| CN113514074B (en) * | 2021-07-09 | 2023-11-17 | 北京航空航天大学 | Intersection path planning method, intersection path planning device and storage medium |
| CN117409000B (en) * | 2023-12-14 | 2024-04-05 | 华能澜沧江水电股份有限公司 | Radar image processing method for slope |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU761072C (en) * | 1998-11-23 | 2003-07-10 | Nestor, Inc. | Traffic light violation prediction and recording system |
| KR101042706B1 (en) * | 2008-11-17 | 2011-06-20 | 허인영 | A distinguishable system for traffic accidents, methods and apparatus thereof |
| CN102280030B (en) * | 2011-07-20 | 2013-09-25 | 杭州海康威视数字技术股份有限公司 | Method and system for detecting vehicle at night |
-
2011
- 2011-12-16 CN CN201110421718.4A patent/CN102496282B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN102496282A (en) | 2012-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102496282B (en) | Traffic intersection signal light state identification method based on RGB color transformation | |
| CN103235938B (en) | The method and system of car plate detection and indentification | |
| CN110197589B (en) | Deep learning-based red light violation detection method | |
| CN103069434B (en) | For the method and system of multi-mode video case index | |
| US9704060B2 (en) | Method for detecting traffic violation | |
| CN105744232B (en) | A kind of method of the transmission line of electricity video external force damage prevention of Behavior-based control analytical technology | |
| CN103824081B (en) | Method for detecting rapid robustness traffic signs on outdoor bad illumination condition | |
| CN102509090B (en) | A kind of vehicle feature recognition device based on public safety video image in sky net engineering | |
| CN104036262B (en) | A kind of method and system of LPR car plates screening identification | |
| Guo et al. | Image-based seat belt detection | |
| CN103955705A (en) | Traffic signal lamp positioning, recognizing and classifying method based on video analysis | |
| Zhang et al. | A multi-feature fusion based traffic light recognition algorithm for intelligent vehicles | |
| CN103761529A (en) | Open fire detection method and system based on multicolor models and rectangular features | |
| CN103488987A (en) | Video-based method and device for detecting traffic lights | |
| CN104715239A (en) | Vehicle color identification method based on defogging processing and weight blocking | |
| CN104732220A (en) | Specific color human body detection method oriented to surveillance videos | |
| CN105205489A (en) | License plate detection method based on color texture analyzer and machine learning | |
| CN103544480A (en) | Vehicle color recognition method | |
| CN105046218A (en) | Multi-feature traffic video smoke detection method based on serial parallel processing | |
| CN112651293A (en) | Video detection method for road illegal stall setting event | |
| CN102306276A (en) | Method for identifying color of vehicle body in video vehicle image based on block clustering | |
| CN112863194B (en) | Image processing method, device, terminal and medium | |
| CN110782409A (en) | Method for removing shadow of multi-motion object | |
| Chen et al. | Traffic congestion classification for nighttime surveillance videos | |
| CN104680195A (en) | Method for automatically recognizing vehicle colors in road intersection video and picture |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201111 Address after: 808, floor 8, building B, business center, gangzhilong science and Technology Park, No. 6, Qinglong Road, Qinghua community, Longhua street, Longhua District, Shenzhen City, Guangdong Province Patentee after: Shenzhen Pengbo Information Technology Co.,Ltd. Address before: 412007 Zhuzhou Research Institute, Hunan University of Technology, 88 West Taishan Road, Tianyuan District, Hunan Patentee before: HUNAN University OF TECHNOLOGY Effective date of registration: 20201111 Address after: Haimen Changle Town Changqing Road 226000 Jiangsu city of Nantong province No. 88 Patentee after: HAIMEN DADE INTELLECTUAL PROPERTY SERVICE Co.,Ltd. Address before: 808, floor 8, building B, business center, gangzhilong science and Technology Park, No. 6, Qinglong Road, Qinghua community, Longhua street, Longhua District, Shenzhen City, Guangdong Province Patentee before: Shenzhen Pengbo Information Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |