CN103458156B - Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions - Google Patents
Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions Download PDFInfo
- Publication number
- CN103458156B CN103458156B CN201310379433.8A CN201310379433A CN103458156B CN 103458156 B CN103458156 B CN 103458156B CN 201310379433 A CN201310379433 A CN 201310379433A CN 103458156 B CN103458156 B CN 103458156B
- Authority
- CN
- China
- Prior art keywords
- video
- measured
- image
- characteristic point
- gray level
- 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.)
- Expired - Fee Related
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions, initially with SIFT feature extraction algorithm, the method judges whether video to be measured exists shake, if there is no shake, then video to be measured is not processed, if there is shake, then remove video to be measured is shaken region, obtain the steady as region of video to be measured；Then the video to be measured after de-jitter is carried out mist elimination process；Advantage is to eliminate the float impact on Vehicle Detection；Eliminate the greasy weather impact on Vehicle Detection, thus under the conditions of the atrocious weather such as strong wind and greasy weather, also can clearly show image, improve detection resolution and the accuracy of detection of video signal.
Description
Technical field
The present invention relates to a kind of traffic incidents detection preprocessing method of video signal, especially relate to a kind of severe weather conditions
Lower traffic incidents detection preprocessing method of video signal.
Background technology
In intelligent transportation event detection system, mostly will have break in traffic rules and regulations detection, and identify and process function intelligent video
The subsystem as intelligent transportation system of detecting system.Now with intelligent video-detect technology be all based on good sky
Vaporous condition builds, and such as fine day, photographic head are completely fixed or daytime is without light interference etc..But, work as monitoring range
Vile weather inside occurs, such as strong wind, dense fog, rains or jamming light source (such as street lamp) etc., now monitor section
The refraction of light, reflection path etc. can occur large change, occur as soon as at reflection to photographic head that fuzzy pictures is unclear, shake or
The phenomenons such as person respective regions is the most dazzling, the existence of these phenomenons makes existing traffic events based on good weather conditions
Detecting system cannot complete monitor task well, under common weather can auto clear detection traffic events, but
Vile weather occurs when, automatically the definition of detection will be poor, does not reaches the effect of detection, occur flase drop,
The problems such as missing inspection.
Summary of the invention
The technical problem to be solved is to provide a kind of detection resolution that can improve video signal and detection essence
Traffic incidents detection preprocessing method of video signal under the severe weather conditions of degree.
The present invention solves the technical scheme that above-mentioned technical problem used: traffic events inspection under a kind of severe weather conditions
Survey preprocessing method of video signal, comprise the following steps:
1. video to be measured is carried out de-jitter: use SIFT feature extraction algorithm judgement video to be measured whether to exist and tremble
Dynamic, if there is no shake, then video to be measured is not processed, if there is shake, then will video to be measured be shaken
Region is removed, and obtains the steady as region of video to be measured；
2. the video to be measured after de-jitter being carried out mist elimination process, detailed process is:
2.-1 contrast threshold is set as λ,A is the most right at the image shot without greasy weather gas of any one width
Ratio angle value, B is any one width total contrast value at the image having greasy weather gas to shoot, and total contrast value is by entering image
After row gradient algorithm, obtain after the absolute value of its all Grad is added；
2.-2 total contrast value asking for video to be measured, if total contrast value of video to be measured is more than or equal to contrast threshold
λ, does not the most carry out mist elimination process, if total contrast value of video to be measured is less than contrast threshold λ, then enters at mist elimination
Reason flow process；
2.-3 methods that have employed rectangular histogram stretching carry out mist elimination process.
Described step 1. in judge video to be measured whether exist the detailed process of shake as:
A., ought as the reference video frames of current video frame using the former frame frame of video of current video frame in video to be measured
Front frame of video and reference video frames are respectively converted into gray level image, and obtain current video frame and the ash of reference video frames respectively
The extreme point in degree graphical rule space:
B. the extreme point of the gray level image metric space of screening current video frame and reference video frames, it is thus achieved that current video frame
The accurate characteristic point with the gray level image of reference video frames；
C. the feature principal direction of the characteristic point of the gray level image of current video frame and reference video frames is calculated；
D. current video is determined according to the feature principal direction of current video frame and the characteristic point of the gray level image of reference video frames
The Feature Descriptor of the characteristic point of the gray level image of frame and reference video frames；
E. carry out Feature Points Matching according to the Feature Descriptor of current video frame and the gray level image of reference video frames, obtain
The characteristic point that in the gray level image of current video frame and reference video frames, the match is successful, if current video frame and reference video
All the match is successful for characteristic point in the gray level image of frame, the most there is not shake, if current video frame and reference video frames
Gray level image in characteristic point to there is coupling unsuccessful, then in video to be measured, retain the characteristic point that the match is successful, go
Except the characteristic point that coupling is unsuccessful.
In described step e, the detailed process of Feature Points Matching is:
E-1. characteristic threshold value is set as R,L represents value less in the length and width of current video frame；
E-2. to each characteristic point in the gray level image of current video frame, all look in the gray level image of reference video frames
Go out the characteristic point that the difference with its Feature Descriptor is minimum, using the two characteristic point as a pair characteristic point, in case coupling；
If e-3. the difference of the distance of a pair characteristic point is less than characteristic threshold value, then it is assumed that this is current video frame to characteristic point
With the same point in reference video frames, this is to Feature Points Matching success；Otherwise, this is unsuccessful to Feature Points Matching.
2. described step have employed the method for rectangular histogram stretching in-3 and carries out the detailed process of mist elimination process and be:
2.-3-1 obtains the R channel image of frame of video, G channel image and the channel B image of video to be measured；
2.-3-2 carries out rectangular histogram stretching to R channel image, and detailed process is:
F. colourity i of pixel in image is represented, i ∈ 0,1 ..., L-1, L are the maximum color of pixel in image
Degree, colourity is total quantity n of the pixel of iiRepresent；
G. the probability that the pixel that colourity in image is i occurs is designated as px(i), thenN represents in image
The total quantity of all pixels；
H. the probability tables that the pixel that colourity in image is i occurs is shown as histogrammic form；
I. p is definedxI the cumulative probability function of () is c (i),C (i) is the accumulative normalization of image
Rectangular histogram；
J. create mapping function y=T (x), use mapping function y=T (x) that c (i) is carried out linearization process, obtain
yi=T (xi)=c (i), c (i) be mapped to [0 ..., 1] territory；
K. formula is usedThe mapping value of c (i) is changed, obtains straight
Image after side's figure stretching, max{x} represents the maximum that x corresponding for i obtains in accumulation histogram, and min{x} represents
Minima in accumulation histogram；
2.-3-3 uses the method for step f～k that G channel image and channel B image are carried out rectangular histogram stretching, obtains G
Image after passage and channel B stretching；
2. the image after R passage, G passage and channel B are stretched by-3-4 merges, and obtains the image after mist elimination processes.
3. described step is additionally provided with step the most afterwards: the video to be measured after being processed by mist elimination carries out light suppression process, tool
Body process is:
3.-1 setting proportion at night as 20%～80%, daytime, proportion was 20%～80%；
3.-2 the video image of video to be measured is converted to gray level image；
3.-3 quantity and all pixels of gray value pixel more than 200 in the gray level image of video images are calculated
The ratio of quantity, if this ratio is positioned at proportion on daytime 20%～80%, video the most to be measured is in daytime, to be measured regards
There is not light interference in Pin, video to be measured is not dealt with, if this ratio is beyond proportion on daytime 20%～80%,
Calculate the quantity of gray value pixel in the range of 0～50 in the gray level image of video image and the quantity of all pixels
Ratio, if this ratio is positioned at proportion at night 20%～80%, video the most to be measured is in night, uses characteristic point
Determine light region with the method for texture information and light region is removed from video to be measured, otherwise, to video to be measured
Do not process.
4. described step is additionally provided with step the most afterwards: the video to be measured after light suppression being processed carries out the rainy day at night and goes to do
Disturbing process, detailed process is:
If 4.-1 video to be measured is in night, in detecting each rule track, whether there is high-brightness region, if do not deposited
, it not the most the rainy day, video to be measured is not processed；If it is present be the rainy day, the entrance rainy day at night goes at interference
Reason flow process；
4.-2 detecting the pixel region of m × m in video image to be measured successively, the span of m is 3～10, if this district
In territory, the pixel value of all pixels is both greater than 200, then it is assumed that this region is the high-brightness region that rainy day illumination causes, from
This high-brightness region is removed by video image to be measured.
Compared with prior art, it is an advantage of the current invention that video to be measured is by using SIFT feature extraction algorithm to judge
No existence shakes, and if there is no shake, does not then process video to be measured, if there is shake, then regards to be measured
Shake region in Pin to remove, obtain the steady as region of video to be measured, eliminate the float impact on Vehicle Detection；And lead to
The method crossing rectangular histogram stretching carries out mist elimination process, eliminates the greasy weather impact on Vehicle Detection, thus in strong wind and greasy weather etc.
Under the conditions of atrocious weather, also can clearly show image, improve detection resolution and the accuracy of detection of video signal；
When video to be measured also carrying out after processing at mist elimination light suppression and processing, light at night can be suppressed detection video
Impact, further increase detection resolution and the accuracy of detection of video signal；
When also video to be measured is carried out after processing in light suppression the rainy day at night go interference to process time, can eliminate the rainy day satisfies the need
The interference of condition detection, further increases detection resolution and the accuracy of detection of video signal.
Detailed description of the invention
Below in conjunction with embodiment, the present invention is described in further detail.
Embodiment: traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions, comprises the following steps:
1. video to be measured is carried out de-jitter: use SIFT (Scale Invariant Feature Transform)
Feature extraction algorithm judges whether video to be measured exists shake, if there is no shake, does not then process video to be measured,
If there is shake, then remove video to be measured is shaken region, obtain the steady as region of video to be measured；
2. the video to be measured after de-jitter being carried out mist elimination process, detailed process is:
2.-1 contrast threshold is set as λ,A is the most right at the image shot without greasy weather gas of any one width
Ratio angle value, B is any one width total contrast value at the image having greasy weather gas to shoot, and total contrast value is by entering image
After row gradient algorithm, obtain after the absolute value of its all Grad is added；
2.-2 total contrast value asking for video to be measured, if total contrast value of video to be measured is more than or equal to contrast threshold
λ, does not the most carry out mist elimination process, if total contrast value of video to be measured is less than contrast threshold λ, then enters at mist elimination
Reason flow process；
2.-3 methods that have employed rectangular histogram stretching carry out mist elimination process.
In the present embodiment, step 1. in judge video to be measured whether exist the detailed process of shake as:
A., ought as the reference video frames of current video frame using the former frame frame of video of current video frame in video to be measured
Front frame of video and reference video frames are respectively converted into gray level image, and obtain current video frame and the ash of reference video frames respectively
The extreme point in degree graphical rule space:
B. the extreme point of the gray level image metric space of screening current video frame and reference video frames, it is thus achieved that current video frame
The accurate characteristic point with the gray level image of reference video frames；
C. the feature principal direction of the characteristic point of the gray level image of current video frame and reference video frames is calculated；
D. current video is determined according to the feature principal direction of current video frame and the characteristic point of the gray level image of reference video frames
The Feature Descriptor of the characteristic point of the gray level image of frame and reference video frames；
E. carry out Feature Points Matching according to the Feature Descriptor of current video frame and the gray level image of reference video frames, obtain
The characteristic point that in the gray level image of current video frame and reference video frames, the match is successful, if current video frame and reference video
All the match is successful for characteristic point in the gray level image of frame, the most there is not shake, if current video frame and reference video frames
Gray level image in characteristic point to there is coupling unsuccessful, then in video to be measured, retain the characteristic point that the match is successful, go
Except the characteristic point that coupling is unsuccessful.
In the present embodiment, in step e, the detailed process of Feature Points Matching is:
E-1. characteristic threshold value is set as R,L represents value less in the length and width of current video frame；
E-2. to each characteristic point in the gray level image of current video frame, all look in the gray level image of reference video frames
Go out the characteristic point that the difference with its Feature Descriptor is minimum, using the two characteristic point as a pair characteristic point, in case coupling；
If e-3. the difference of the distance of a pair characteristic point is less than characteristic threshold value, then it is assumed that this is current video frame to characteristic point
With the same point in reference video frames, this is to Feature Points Matching success；Otherwise, this is unsuccessful to Feature Points Matching.
In the present embodiment, 2. step have employed the method for rectangular histogram stretching in-3 and carries out the detailed process of mist elimination process and be:
2.-3-1 obtains the R channel image of frame of video, G channel image and the channel B image of video to be measured；
2.-3-2 carries out rectangular histogram stretching to R channel image, and detailed process is:
F. colourity i of pixel in image is represented, i ∈ 0,1 ..., L-1, L are the maximum color of pixel in image
Degree, colourity is total quantity n of the pixel of iiRepresent；
G. the probability that the pixel that colourity in image is i occurs is designated as px(i), thenN represents in image
The total quantity of all pixels；
H. the probability tables that the pixel that colourity in image is i occurs is shown as histogrammic form；
I. p is definedxI the cumulative probability function of () is c (i),C (i) is the accumulative normalization of image
Rectangular histogram；
J. create mapping function y=T (x), use mapping function y=T (x) that c (i) is carried out linearization process, obtain
yi=T (xi)=c (i), c (i) be mapped to [0 ..., 1] territory；
K. formula is usedThe mapping value of c (i) is changed, obtains straight
Image after side's figure stretching, max{x} represents the maximum that x corresponding for i obtains in accumulation histogram, and min{x} represents
Minima in accumulation histogram；
2.-3-3 uses the method for step f～k that G channel image and channel B image are carried out rectangular histogram stretching, obtains G
Image after passage and channel B stretching；
2. the image after R passage, G passage and channel B are stretched by-3-4 merges, and obtains the image after mist elimination processes.
In the present embodiment, 3. step is additionally provided with step the most afterwards: the video to be measured after being processed by mist elimination is carried out at light suppression
Reason, detailed process is:
3.-1 setting proportion at night as 20%～80%, daytime, proportion was 20%～80%；
3.-2 the video image of video to be measured is converted to gray level image；
3.-3 quantity and all pixels of gray value pixel more than 200 in the gray level image of video images are calculated
The ratio of quantity, if this ratio is positioned at proportion on daytime 20%～80%, video the most to be measured is in daytime, to be measured regards
There is not light interference in Pin, video to be measured is not dealt with, if this ratio is beyond proportion on daytime 20%～80%,
Calculate the quantity of gray value pixel in the range of 0～50 in the gray level image of video image and the quantity of all pixels
Ratio, if this ratio is positioned at proportion at night 20%～80%, video the most to be measured is in night, uses characteristic point
Determine light region with the method for texture information and light region is removed from video to be measured, otherwise, to video to be measured
Do not process.
In the present embodiment, 4. step is additionally provided with step the most afterwards: the video to be measured after light suppression being processed carries out rain at night
It goes interference to process, and detailed process is:
If 4.-1 video to be measured is in night, in detecting each rule track, whether there is high-brightness region, if do not deposited
, it not the most the rainy day, video to be measured is not processed；If it is present be the rainy day, the entrance rainy day at night goes at interference
Reason flow process；
4.-2 detecting the pixel region of m × m in video image to be measured successively, the span of m is 3～10, if this district
In territory, the pixel value of all pixels is both greater than 200, then it is assumed that this region is the high-brightness region that rainy day illumination causes, from
This high-brightness region is removed by video image to be measured.
Claims (5)
1. traffic incidents detection preprocessing method of video signal under a severe weather conditions, it is characterised in that include following
Step:
1. video to be measured is carried out de-jitter: use SIFT feature extraction algorithm judgement video to be measured whether to exist and tremble
Dynamic, if there is no shake, then video to be measured is not processed, if there is shake, then will video to be measured be shaken
Region is removed, and obtains the steady as region of video to be measured；
2. the video to be measured after de-jitter being carried out mist elimination process, detailed process is:
2.-1 contrast threshold is set as λ,A is the most right at the image shot without greasy weather gas of any one width
Ratio angle value, B is any one width total contrast value at the image having greasy weather gas to shoot, and total contrast value is by entering image
After row gradient algorithm, obtain after the absolute value of its all Grad is added；
2.-2 total contrast value asking for video to be measured, if total contrast value of video to be measured is more than or equal to contrast threshold
λ, does not the most carry out mist elimination process, if total contrast value of video to be measured is less than contrast threshold λ, then enters at mist elimination
Reason flow process；
2.-3 methods that have employed rectangular histogram stretching carry out mist elimination process；
2. described step have employed the method for rectangular histogram stretching in-3 and carries out the detailed process of mist elimination process and be:
2.-3-1 obtains the R channel image of frame of video, G channel image and the channel B image of video to be measured；
2.-3-2 carries out rectangular histogram stretching to R channel image, and detailed process is:
F. colourity i of pixel in image is represented, i ∈ 0,1 ..., L-1, L are the maximum color of pixel in image
Degree, colourity is total quantity n of the pixel of iiRepresent；
G. the probability that the pixel that colourity in image is i occurs is designated as px(i), thenN represents in image
The total quantity of all pixels；
H. the probability tables that the pixel that colourity in image is i occurs is shown as histogrammic form；
I. p is definedxI the cumulative probability function of () is c (i),C (i) is the accumulative normalization of image
Rectangular histogram；
J. create mapping function y=T (x), use mapping function y=T (x) that c (i) is carried out linearization process, obtain
yi=T (xi)=c (i), c (i) be mapped to [0 ..., 1] territory；
K. formula y ' is usedi=yi(max{x}-min{x}) mapping value of c (i) is changed by+min{x}, obtains straight
Image after side's figure stretching, max{x} represents the maximum that x corresponding for i obtains in accumulation histogram, and min{x} represents
Minima in accumulation histogram；
2.-3-3 uses the method for step f～k that G channel image and channel B image are carried out rectangular histogram stretching, obtains G
Image after passage and channel B stretching；
2. the image after R passage, G passage and channel B are stretched by-3-4 merges, and obtains the image after mist elimination processes.
Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions the most according to claim 1,
It is characterized in that judging during described step is 1. video to be measured whether exist the detailed process of shake as:
A., ought as the reference video frames of current video frame using the former frame frame of video of current video frame in video to be measured
Front frame of video and reference video frames are respectively converted into gray level image, and obtain current video frame and the ash of reference video frames respectively
The extreme point in degree graphical rule space:
B. the extreme point of the gray level image metric space of screening current video frame and reference video frames, it is thus achieved that current video frame
The accurate characteristic point with the gray level image of reference video frames；
C. the feature principal direction of the characteristic point of the gray level image of current video frame and reference video frames is calculated；
D. current video is determined according to the feature principal direction of current video frame and the characteristic point of the gray level image of reference video frames
The Feature Descriptor of the characteristic point of the gray level image of frame and reference video frames；
E. carry out Feature Points Matching according to the Feature Descriptor of current video frame and the gray level image of reference video frames, obtain
The characteristic point that in the gray level image of current video frame and reference video frames, the match is successful, if current video frame and reference video
All the match is successful for characteristic point in the gray level image of frame, the most there is not shake, if current video frame and reference video frames
Gray level image in characteristic point to there is coupling unsuccessful, then in video to be measured, retain the characteristic point that the match is successful, go
Except the characteristic point that coupling is unsuccessful.
Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions the most according to claim 2,
It is characterized in that in described step e, the detailed process of Feature Points Matching is:
E-1. characteristic threshold value is set as R,L represents value less in the length and width of current video frame；
E-2. to each characteristic point in the gray level image of current video frame, all look in the gray level image of reference video frames
Go out the characteristic point that the difference with its Feature Descriptor is minimum, using the two characteristic point as a pair characteristic point, in case coupling；
If e-3. the difference of the distance of a pair characteristic point is less than characteristic threshold value, then it is assumed that this is current video frame to characteristic point
With the same point in reference video frames, this is to Feature Points Matching success；Otherwise, this is unsuccessful to Feature Points Matching.
Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions the most according to claim 1,
It is characterized in that 3. described step is additionally provided with step the most afterwards: the video to be measured after being processed by mist elimination is carried out at light suppression
Reason, detailed process is:
3.-1 setting proportion at night as 20%～80%, daytime, proportion was 20%～80%；
3.-2 the video image of video to be measured is converted to gray level image；
3.-3 quantity and all pixels of gray value pixel more than 200 in the gray level image of video images are calculated
The ratio of quantity, if this ratio is positioned at proportion on daytime 20%～80%, video the most to be measured is in daytime, to be measured regards
There is not light interference in Pin, video to be measured is not dealt with, if this ratio is beyond proportion on daytime 20%～80%,
Calculate the quantity of gray value pixel in the range of 0～50 in the gray level image of video image and the quantity of all pixels
Ratio, if this ratio is positioned at proportion at night 20%～80%, video the most to be measured is in night, uses characteristic point
Determine light region with the method for texture information and light region is removed from video to be measured, otherwise, to video to be measured
Do not process.
Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions the most according to claim 4,
It is characterized in that 4. described step is additionally provided with step the most afterwards: the video to be measured after light suppression being processed carries out rain at night
It goes interference to process, and detailed process is:
If 4.-1 video to be measured is in night, in detecting each rule track, whether there is high-brightness region, if do not deposited
, it not the most the rainy day, video to be measured is not processed；If it is present be the rainy day, the entrance rainy day at night goes at interference
Reason flow process；
4.-2 detecting the pixel region of m × m in video image to be measured successively, the span of m is 3～10, if this district
In territory, the pixel value of all pixels is both greater than 200, then it is assumed that this region is the high-brightness region that rainy day illumination causes, from
This high-brightness region is removed by video image to be measured.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310379433.8A CN103458156B (en) | 2013-08-27 | 2013-08-27 | Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310379433.8A CN103458156B (en) | 2013-08-27 | 2013-08-27 | Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103458156A CN103458156A (en) | 2013-12-18 |
| CN103458156B true CN103458156B (en) | 2016-08-10 |
Family
ID=49740080
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310379433.8A Expired - Fee Related CN103458156B (en) | 2013-08-27 | 2013-08-27 | Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103458156B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10870398B2 (en) * | 2015-07-28 | 2020-12-22 | Ford Global Technologies, Llc | Vehicle with hyperlapse video and social networking |
| CN105678274B (en) * | 2016-01-13 | 2019-04-02 | 符锌砂 | Monitoring Identify Environment based on characteristics of image under a kind of rainy day environment |
| CN105574830B (en) * | 2016-02-04 | 2020-02-21 | 沈阳工业大学 | Low-quality image enhancement method under extreme weather condition |
| CN105976330B (en) * | 2016-04-27 | 2019-04-09 | 大连理工大学 | A kind of embedded greasy weather real time video image stabilization |
| CN108446657B (en) | 2018-03-28 | 2022-02-25 | 京东方科技集团股份有限公司 | Gesture jitter recognition method and device and gesture recognition method |
| CN109636793A (en) * | 2018-12-14 | 2019-04-16 | 中航华东光电（上海）有限公司 | The detection system and its detection method of display |
| CN109493613A (en) * | 2018-12-27 | 2019-03-19 | 江苏中科智能系统有限公司 | Traffic events monitor system |
| CN111754783A (en) * | 2020-06-19 | 2020-10-09 | 张永双 | Illegal parking warning and snapshot method and system based on cloud computing and artificial intelligence |
| CN113255612A (en) * | 2021-07-05 | 2021-08-13 | 智道网联科技(北京)有限公司 | Preceding vehicle starting reminding method and system, electronic device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819286A (en) * | 2010-04-09 | 2010-09-01 | 东南大学 | Image grey level histogram-based foggy day detection method |
| CN102170574A (en) * | 2011-05-23 | 2011-08-31 | 北京工业大学 | Real-time video defogging system |
| CN103021177A (en) * | 2012-11-05 | 2013-04-03 | 北京理工大学 | Method and system for processing traffic monitoring video image in foggy day |
-
2013
- 2013-08-27 CN CN201310379433.8A patent/CN103458156B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819286A (en) * | 2010-04-09 | 2010-09-01 | 东南大学 | Image grey level histogram-based foggy day detection method |
| CN102170574A (en) * | 2011-05-23 | 2011-08-31 | 北京工业大学 | Real-time video defogging system |
| CN103021177A (en) * | 2012-11-05 | 2013-04-03 | 北京理工大学 | Method and system for processing traffic monitoring video image in foggy day |
Non-Patent Citations (2)
| Title |
|---|
| 基于暗原色和直方图匹配的雾天图像增强算法;张洪坤等;《计算机工程》;20120131;第38卷(第1期);全文 * |
| 基于黑色通道的图像快速去雾优化算法;褚宏莉等;《电子学报》;20130430;第41卷(第4期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103458156A (en) | 2013-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103458156B (en) | Traffic incidents detection preprocessing method of video signal under a kind of severe weather conditions | |
| CN110555361B (en) | Image processing method for lane classification | |
| US9225943B2 (en) | PTZ video visibility detection method based on luminance characteristic | |
| CN104952060B (en) | A kind of infrared pedestrian's area-of-interest adaptivenon-uniform sampling extracting method | |
| Gallen et al. | Towards night fog detection through use of in-vehicle multipurpose cameras | |
| CN110415544B (en) | Disaster weather early warning method and automobile AR-HUD system | |
| CN111832536A (en) | Lane line detection method and device | |
| TW201716266A (en) | Image inpainting system area and method using the same | |
| CN104156727B (en) | Lamplight inverted image detection method based on monocular vision | |
| CN110088766A (en) | Lane detection method, Lane detection device and non-volatile memory medium | |
| CN111860120A (en) | Automatic shielding detection method and device for vehicle-mounted camera | |
| CN110929676A (en) | Deep learning-based real-time detection method for illegal turning around | |
| KR101026778B1 (en) | Vehicle image detection apparatus | |
| CN109191492B (en) | Intelligent video black smoke vehicle detection method based on contour analysis | |
| CN109919062A (en) | A kind of road scene weather recognition methods based on characteristic quantity fusion | |
| US8229170B2 (en) | Method and system for detecting a signal structure from a moving video platform | |
| KR101795652B1 (en) | Device for Measuring Visibility for Fog Guardian Device | |
| CN109766846B (en) | Video-based self-adaptive multi-lane traffic flow detection method and system | |
| Skodras et al. | Rear lights vehicle detection for collision avoidance | |
| JP2004086417A (en) | Method and device for detecting pedestrian on zebra crossing | |
| WO2008088409A2 (en) | Real-time dynamic content based vehicle tracking, traffic monitoring, and classification system | |
| Zong et al. | Traffic light detection based on multi-feature segmentation and online selecting scheme | |
| Yoshimori et al. | License plate detection using hereditary threshold determine method | |
| Li et al. | A fog level detection method based on grayscale features | |
| JP2022151740A (en) | Method for determining images with possibility of having false negative object detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160810 Termination date: 20190827 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |