CN104537672A - Road environment monotony detection method based on videos - Google Patents

Road environment monotony detection method based on videos Download PDF

Info

Publication number
CN104537672A
CN104537672A CN201510003334.9A CN201510003334A CN104537672A CN 104537672 A CN104537672 A CN 104537672A CN 201510003334 A CN201510003334 A CN 201510003334A CN 104537672 A CN104537672 A CN 104537672A
Authority
CN
China
Prior art keywords
gray
value
scale
road environment
road
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.)
Granted
Application number
CN201510003334.9A
Other languages
Chinese (zh)
Other versions
CN104537672B (en
Inventor
徐婷
何立明
尚文科
杨新新
陈刚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Changan University
Original Assignee
Changan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Changan University filed Critical Changan University
Priority to CN201510003334.9A priority Critical patent/CN104537672B/en
Publication of CN104537672A publication Critical patent/CN104537672A/en
Application granted granted Critical
Publication of CN104537672B publication Critical patent/CN104537672B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/77—Determining position or orientation of objects or cameras using statistical methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30248—Vehicle exterior or interior
    • G06T2207/30252—Vehicle exterior; Vicinity of vehicle
    • G06T2207/30256—Lane; Road marking

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a road environment monotony detection method based on videos. The method includes the following steps that the road environment videos of roads to be detected are acquired, and color images in a video image frame sequence are converted into grey-scale maps; grey-scale values of all pixel points of each grey-scale map are calculated and stored in a pixel point grey-scale value matrix Ai; a grey-scale value difference matrix Mi of every two adjacent grey-scale maps is calculated; a first threshold value A is set, the number Si of distinguished elements of elements with the values larger than the first threshold value A in the grey-scale value difference matrixes Mi is counted; a second threshold value B is set, when the number Si of the distinguished elements is larger than the second threshold value B, road environments corresponding to Ai and Ai+1 are regarded to have no monotony, and when the number Si of the distinguished elements is smaller than or equal to the second threshold value B, the road environments corresponding to Ai and Ai+1 are regarded to have monotony; a corresponding map of the grey-scale maps and roads is established, actual road positions corresponding to Ai and Ai+1 are found, and then whether the actual road environments have monotony or not can be known.

Description

A kind of detection method of the road environment monotonicity based on video
Technical field
The invention belongs to traffic safety technical field, specifically disclose a kind of detection method of the road environment monotonicity based on video.
Background technology
Along with China's expanding economy, vehicle guaranteeding organic quantity improves year by year, but traffic safety situation allows of no optimist, and the traffic hazard caused because of driver fatigue gets more and more.Driver's action repeatedly, makes its psychology, certain change occurs physiology, occur dispersion attention, doze, the visual field narrow, information leaks phenomenons such as seeing.Therefore, fatigue driving has a strong impact on traffic safety, becomes the major hidden danger of traffic hazard.Cause the reason of fatigue driving a lot, one of wherein important reason is exactly the unicity of road landscape.During the highway up train that driver is straight on road surface, environment is dull, quick, dull, the high operation repeated can make driver's vision, mental fatigue.
Sentencing method for distinguishing for environment monotonicity is by train experiment substantially, and eye tracker device and electrocardio instrument gather the parameter of driver's physiology, psychology, judge that whether driver is tired by physiology, psychological parameter.At present also not based on the view unicity method of discrimination of video sequence.
Summary of the invention
For problems of the prior art, the object of this invention is to provide a kind of detection method of the road environment monotonicity based on video, the monotonicity of road environment is made accurately, stablizes and judge reliably, and for road construction planning and road landscape construction provide support, thus the traffic hazard reduced because fatigue driving causes, improve road safety level.
In order to achieve the above object, technical scheme of the present invention realizes like this.
Based on a detection method for the road environment monotonicity of video, it is characterized in that, comprise the following steps:
Step one, vehicle normally travels on road to be detected, is gathered the road environment video of road to be detected by the video capture device on vehicle;
Step 2, be the video image frame sequence of m × n-pixel size by the road environment Video processing collected, and the coloured image in video image frame sequence is converted to gray-scale map, m, n are natural number;
Step 3, calculates the gray-scale value of all pixels of each frame gray-scale map, saves as pixel gray-scale value matrix A i, A ifor the matrix of m × n, represent the pixel gray-scale value matrix of the i-th frame gray-scale map, i is natural number;
Step 4, calculates the pixel gray value differences value matrix M of adjacent two frame gray-scale maps i, M ifor the matrix of m × n, M i=A i+1-A i, i is natural number;
Step 5, the value of setting first threshold A, statistical pixel point gray value differences value matrix M iin be greater than the number of the element of first threshold A, be designated as distinguished element number S i, i is natural number;
Step 6, resets the value of Second Threshold B, compares distinguished element number S iwith the size of Second Threshold B:
As distinguished element number S iduring > Second Threshold B, namely think A i, A i+1corresponding road environment does not have monotonicity;
As distinguished element number S iduring≤Second Threshold B, namely think A i, A i+1corresponding road environment has monotonicity;
Step 7, sets up gray-scale map and the corresponding table of road, finds A i, A i+1corresponding real road position, can know whether this section of real road environment has monotonicity.
Through above-mentioned steps, can judge whether road environment has monotonicity, detect complete.
Feature of the present invention is:
(1) computing formula of gray-scale value described in step 3 is: Gray=0.299 × R+0.587 × G+0.114 × B, and in formula, Gray is gray-scale value, and R, G, B are respectively the three primary colors numerical value of this pixel.
(2) pixel gray value differences value matrix M described in step 4 icomputing method adopt frame difference method.
(3) value of first threshold A described in step 5 is the mean value of the absolute value of the pixel gray-scale value difference of adjacent two frame gray-scale maps.
(4) value of Second Threshold B described in step 6 adopts formula B=η × m × n, and in formula, η is experience factor, between between.
Compared with prior art, the present invention has the following advantages and useful technique effect.The detection method of road environment monotonicity of the present invention is mainly used in traffic monitoring department and road construction department to the judgement of road environment, can meet the requirement of the accuracy to road environment data, reliability and accuracy of detection.The present invention is compared with existing road environment detection method, do not need by compressing video, thus avoid extra coding, a large amount of data operations can be saved, reduce the requirement to hardware, monotonicity detection can be carried out to all environment in road visual range again simultaneously, and simple and reliable, be easy to realize, have broad application prospects.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the detection method of road environment monotonicity based on video;
Fig. 2 is the gray-scale map of G107 Hebei-Zhuozhou section No. 134 road environment frame of video;
Fig. 3 is the gray-scale map of G107 Hebei-Zhuozhou section No. 135 road environment frame of video;
Fig. 4 is the gray-scale map of G107 Hebei-Zhuozhou section No. 597 road environment frame of video;
Fig. 5 is the gray-scale map of G107 Hebei-Zhuozhou section No. 598 road environment frame of video;
Fig. 6 is distinguished element number S 134statistics figure;
Fig. 7 is distinguished element number S 597statistics figure.
Embodiment
Below in conjunction with specific embodiment, the present invention is described in further details, but the present invention is not limited thereto embodiment.
The section to be detected of the present embodiment is G107 Hebei-Zhuozhou section, carries out monotonicity inspection by the road environment monotonicity detection method based on video of the present invention to G107 Hebei-Zhuozhou section.With reference to Fig. 1, it is the process flow diagram of the detection method of the road environment monotonicity based on video.
Specifically comprise the following steps:
Step one, gathers one section of road environment video by the culminant star ZX-GN8600 high-definition camera on vehicle in G107 high speed Hebei-section, Zhuozhou.
Step 2, be the video image frame sequence of 1080 × 1902 pixel sizes with the frequency processing of 2 frames/second by the road environment video collected, and the gray-scale map coloured image in video image frame sequence is converted to only containing pixel gray-scale value, with reference to Fig. 2-Fig. 5.
Step 3, calculate the pixel gray-scale value of all pixels of each frame gray-scale map, computing formula is: Gray=0.299 × R+0.587 × G+0.114 × B, and in formula, Gray is gray-scale value, and R, G, B are respectively the three primary colors numerical value of this pixel.Save as the pixel gray-scale value matrix A of each frame gray-scale map i, A ibe the matrix of 1080 × 1902, represent the pixel gray-scale value matrix of the i-th frame gray-scale map, i is natural number, i=1,2 ..., 3600.Gray-scale value matrix A isuch as formula 1, gray-scale value matrix A i+1such as formula 2.
A i = A 11 A 12 A 13 . . . . . . A 11900 A 11901 A 11902 A 21 A 22 A 23 . . . . . . A 21900 A 21901 A 21902 A 31 A 32 A 33 . . . . . . A 31900 A 31901 A 31902 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A 10781 A 10782 A 10783 . . . . . . A 10781900 A 10781901 A 10781902 A 10791 A 10792 A 10793 . . . . . . A 10791900 A 10791901 A 10791902 A 10801 A 10802 A 10803 . . . . . . A 10801900 A 10801900 A 10801902 (formula 1)
A i + 1 = B 11 B 12 B 13 . . . . . . B 11900 B 11901 B 11902 B 21 B 22 B 23 . . . . . . B 21900 B 21901 B 21902 B 31 B 32 B 33 . . . . . . B 31900 B 31901 B 31902 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B 10781 B 10782 B 10783 . . . . . . B 10781900 B 10781901 B 10781902 B 10791 B 10792 B 10793 . . . . . . B 10791900 B 10791901 B 10791902 B 10801 B 10802 B 10803 . . . . . . B 10801900 B 10801900 B 10801902 (formula 2)
Formula 1 represents the pixel gray-scale value matrix A of the i-th frame gray-scale map i, formula 2 represents the pixel gray-scale value matrix A of the i-th+1 frame gray-scale map i+1; Pixel value size due to the present embodiment image is 1080 × 1902, therefore its gray-scale value entry of a matrix element number is also 1080 × 1902, and namely matrix has 1080 row, 1902 row.
Step 4, utilizes frame difference method, calculates the pixel gray value differences value matrix M of adjacent two frame gray-scale maps i, M ibe the matrix of 1080 × 1902, M i=A i+1-A i, i is natural number, i=1,2 ..., 3599; Pixel gray value differences value matrix M isuch as formula 3.
M i = M 11 M 12 M 13 . . . . . . M 11900 M 11901 M 11902 M 21 M 22 M 23 . . . . . . M 21900 M 21901 M 21902 M 31 M 32 M 33 . . . . . . M 31900 M 31901 M 31902 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . M 10781 M 10782 M 10783 . . . . . . M 10781900 M 10781901 M 10781902 M 10791 M 10792 M 10793 . . . . . . M 10791900 M 10791901 M 10791902 M 10801 M 10802 M 10803 . . . . . . M 10801900 M 10801900 M 10801902 (formula 3)
Step 5, the value of setting first threshold A, the value of the present embodiment first threshold A is the mean value of the absolute value of the pixel gray-scale value difference of adjacent two frame gray-scale maps, i.e. first threshold A=50.Statistics gray value differences value matrix M iin be greater than the number of the element of 50, be designated as distinguished element number S i, i is natural number, i=1,2 ..., 3599.Wherein, pixel gray value differences value matrix M 134in be greater than the element of 50 number be 13048, i.e. distinguished element number S 134=13048; Pixel gray value differences value matrix M 597in be greater than the element of 50 number be 165691, i.e. distinguished element number S 597=165691.With reference to Fig. 6, Fig. 7.
The value of first threshold A can also rule of thumb value, as judged that it has the video frame image of monotonicity by visual inspection by first selected part, by the anti-value pushing back first threshold A of step 5.
Step 6, the value of setting Second Threshold B, wherein B=η × m × n, in formula, η is experience factor, and the present embodiment η gets m × n=1080 × 1902, calculate Second Threshold B=50000, compare distinguished element number S iwith 50000 size.Wherein, distinguished element number S 134< 50000, then A 134, A 135corresponding real road environment has monotonicity; Distinguished element number S 597> 50000, then A 597, A 598corresponding real road environment does not have monotonicity.
The value of Second Threshold B can also rule of thumb value, as can be determined the value of B according to the size of m, n, also can judge that it has the video frame image of monotonicity by visual inspection by first selected part, then by the anti-value pushing back Second Threshold B of step 6.
Step 7, sets up gray-scale map and the corresponding table of road, finds A iand A i+1corresponding real road position, can know whether this section of real road has monotonicity.Lookup result shows, G107 Hebei-No. 134-135, Zhuozhou section this section of road environment has monotonicity, and G107 Hebei-No. 597-598, Zhuozhou section this section of road environment does not have monotonicity.
Visual inspection detects the gray-scale map of G107 Hebei-Zhuozhou section No. 134 road environment frame of video and the gray-scale map of G107 Hebei-Zhuozhou section No. 135 road environment frame of video, judges that No. 134-135 this section of road environment has monotonicity; Visual inspection detects the gray-scale map of G107 Hebei-Zhuozhou section No. 597 road environment frame of video and the gray-scale map of G107 Hebei-Zhuozhou section No. 598 road environment frame of video, judges that No. 597-598 this section of road environment does not have monotonicity.Consistent with step 6 testing result, detection method accurate and effective of the present invention is described.
If inconsistent with step 6 testing result, the size of first threshold A and Second Threshold B need be adjusted, until two testing results are consistent, and with the first threshold A after adjustment and Second Threshold B, step 5 and step 6 are repeated to all gray-scale maps, obtain testing result accurately.
Although be described embodiment of the present invention above, the present invention is not limited to above-mentioned specific embodiments, and above-mentioned specific embodiments is only schematic, guiding, instead of restrictive.Those of ordinary skill in the art is under the enlightenment of this instructions, and when not departing from the scope that the claims in the present invention are protected, can also make a variety of forms, these all belong to the row of the present invention's protection.

Claims (5)

1., based on a detection method for the road environment monotonicity of video, it is characterized in that, comprise the following steps:
Step one, vehicle normally travels on road to be detected, is gathered the road environment video of road to be detected by the video capture device on vehicle;
Step 2, be the video image frame sequence of m × n-pixel size by the road environment Video processing collected, and the coloured image in video image frame sequence is converted to gray-scale map, m, n are natural number;
Step 3, calculates the gray-scale value of all pixels of each frame gray-scale map, saves as pixel gray-scale value matrix A i, A ifor the matrix of m × n, represent the pixel gray-scale value matrix of the i-th frame gray-scale map, i is natural number;
Step 4, calculates the pixel gray value differences value matrix M of adjacent two frame gray-scale maps i, M ifor the matrix of m × n, M i=A i+1-A i, i is natural number;
Step 5, the value of setting first threshold A, statistical pixel point gray value differences value matrix M iin be greater than the number of the element of first threshold A, be designated as distinguished element number S i, i is natural number;
Step 6, resets the value of Second Threshold B, compares distinguished element number S iwith the size of Second Threshold B:
As distinguished element number Si > Second Threshold B, namely think A i, A i+1corresponding road environment does not have monotonicity;
As distinguished element number Si≤Second Threshold B, namely think A i, A i+1corresponding road environment has monotonicity;
Step 7, sets up gray-scale map and the corresponding table of road, finds A i, A i+1corresponding real road position, can know whether this section of real road environment has monotonicity.
2. the detection method of the road environment monotonicity based on video according to claim 1, it is characterized in that, the computing formula of gray-scale value described in step 3 is: Gray=0.299 × R+0.587 × G+0.114 × B, in formula, Gray is gray-scale value, and R, G, B are respectively the three primary colors numerical value of this pixel.
3. the detection method of the road environment monotonicity based on video according to claim 1, is characterized in that, pixel gray value differences value matrix M described in step 4 icomputing method adopt frame difference method.
4. the detection method of the road environment monotonicity based on video according to claim 1, is characterized in that, the value of first threshold A described in step 5 is the mean value of the absolute value of the pixel gray-scale value difference of adjacent two frame gray-scale maps.
5. the detection method of the road environment monotonicity based on video according to claim 1, is characterized in that, the value of Second Threshold B described in step 6 adopts formula B=η × m × n, and in formula, η is experience factor, between between.
CN201510003334.9A 2015-01-05 2015-01-05 A kind of detection method of the road environment monotonicity based on video Expired - Fee Related CN104537672B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510003334.9A CN104537672B (en) 2015-01-05 2015-01-05 A kind of detection method of the road environment monotonicity based on video

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510003334.9A CN104537672B (en) 2015-01-05 2015-01-05 A kind of detection method of the road environment monotonicity based on video

Publications (2)

Publication Number Publication Date
CN104537672A true CN104537672A (en) 2015-04-22
CN104537672B CN104537672B (en) 2018-11-30

Family

ID=52853191

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510003334.9A Expired - Fee Related CN104537672B (en) 2015-01-05 2015-01-05 A kind of detection method of the road environment monotonicity based on video

Country Status (1)

Country Link
CN (1) CN104537672B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1506686A (en) * 2002-12-09 2004-06-23 北京中星微电子有限公司 Moving image detecting method
US20050219057A1 (en) * 2004-03-30 2005-10-06 Mitsubishi Fuso Truck And Bus Corporation Consciousness judging apparatus
CN101699469A (en) * 2009-11-09 2010-04-28 南京邮电大学 Method for automatically identifying action of writing on blackboard of teacher in class video recording
JP2013061270A (en) * 2011-09-14 2013-04-04 Mitsubishi Electric Corp Present position calculation device and present position calculation method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1506686A (en) * 2002-12-09 2004-06-23 北京中星微电子有限公司 Moving image detecting method
US20050219057A1 (en) * 2004-03-30 2005-10-06 Mitsubishi Fuso Truck And Bus Corporation Consciousness judging apparatus
CN101699469A (en) * 2009-11-09 2010-04-28 南京邮电大学 Method for automatically identifying action of writing on blackboard of teacher in class video recording
JP2013061270A (en) * 2011-09-14 2013-04-04 Mitsubishi Electric Corp Present position calculation device and present position calculation method

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
PIERRE THIFFAULT ET AL: "Monotony of road environment and driver fatigue:a similator study", 《ACCIDENT ANALYSIS AND PREVENTION》 *
毛科俊: "道路环境单调性对驾驶疲劳的影响机理及对策研究", 《中国博士学位论文全文数据库 工程科技II辑》 *

Also Published As

Publication number Publication date
CN104537672B (en) 2018-11-30

Similar Documents

Publication Publication Date Title
CN101430195B (en) Method for computing electric power line ice-covering thickness by using video image processing technology
CN105424655B (en) A kind of visibility detecting method based on video image
CN105512623B (en) Based on multisensor travelling in fog day vision enhancement and visibility early warning system and method
CN103902976B (en) A kind of pedestrian detection method based on infrared image
CN105261018B (en) Visibility detecting method based on optical model and dark primary priori theoretical
CN104634784B (en) atmospheric visibility monitoring method and device
CN102509102A (en) Visibility measuring method based on image study
CN103150908B (en) Average vehicle speed detecting method based on video
CN105354806A (en) Dark channel based rapid defogging method and system
CN106127715A (en) A kind of image defogging method and system
CN104240239A (en) Method for detecting local road segment hazy weather based on road image
CN102779349A (en) Foggy day detecting method based on image color spatial feature
CN104851288B (en) Traffic light positioning method
CN109190483A (en) A kind of method for detecting lane lines of view-based access control model
CN109948433A (en) A kind of embedded human face tracing method and device
CN103514460B (en) Video monitoring multi-view-angle vehicle detecting method and device
CN106408526A (en) Visibility detection method based on multilayer vectogram
CN103149603B (en) Road weather detection method based on video
CN104598914A (en) Skin color detecting method and device
CN104766337A (en) Aircraft landing vision enhancement method based on runway boundary enhancement
Wongsaree et al. Distance detection technique using enhancing inverse perspective mapping
CN103927523B (en) Fog level detection method based on longitudinal gray features
CN116989679A (en) Highway high slope displacement integrated monitoring method based on image processing
CN115994901A (en) Automatic road disease detection method and system
CN106485697A (en) A kind of roadbed subsidence based on binocular vision and foreign matter detecting method

Legal Events

Date Code Title Description
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20181130

Termination date: 20210105

CF01 Termination of patent right due to non-payment of annual fee