CN101098462A - Chroma deviation and brightness deviation combined video moving object detection method - Google Patents
Chroma deviation and brightness deviation combined video moving object detection method Download PDFInfo
- Publication number
- CN101098462A CN101098462A CNA2007100437348A CN200710043734A CN101098462A CN 101098462 A CN101098462 A CN 101098462A CN A2007100437348 A CNA2007100437348 A CN A2007100437348A CN 200710043734 A CN200710043734 A CN 200710043734A CN 101098462 A CN101098462 A CN 101098462A
- Authority
- CN
- China
- Prior art keywords
- pixel
- difference
- value
- background
- threshold value
- 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
Links
Abstract
A video mobile target check method combining chromaticity bias and brightness bias, belonging to video processing technical field, comprises that first using statistic method to extract a background frame via different channels, then using middle-value filter to eliminate the interface of color imbalance point in the background frame, at last, replacing former RGB three-channel chromaticity value at saturation jump point into chromaticity average value of RGB three channels, before using brightness difference and chromaticity difference information to check foreground object, first filtering the brightness difference and chromaticity difference matrix, using static Gauss mode to calculate out the threshold value of the brightness difference, and finding the threshold value of the chromaticity difference according to experience value, at last, holding the boundary of checked result. The invention can accurately extract background frame without foreground information, under less video frames and slow foreground motion, which can accurately calculate out the brightness threshold value of foreground object according to the characters of different video sequences and store the boundary of mobile target, thereby accurately checking the mobile target in video sequence.
Description
Technical field
What the present invention relates to is a kind of moving target detecting method of video, and particularly a kind of video moving object detection method in conjunction with chromaticity distortion and luminance deviation belongs to technical field of video processing.
Background technology
The moving object detection of video detects moving object exactly from video sequence, be the important research content of applications such as computer vision, video image tracking.Can moving target be detected accurately and effectively and be extracted, and directly has influence on the treatment effect that follow-up target classification, tracking and behavior are understood.
Moving target detecting method commonly used at present has optical flow method, frame-to-frame differences point-score and background subtraction point-score, and these three kinds of methods all have pluses and minuses separately.Wherein the optical flow method testing result is the most accurate, but operational formula complexity, amount of calculation are big, real-time is poor, to the hardware requirement height.The frame-to-frame differences point-score utilizes adjacent image change in information amount, distinguish background and foreground moving target, advantage is that amount of calculation is littler than optical flow method, variation to scene light is not too responsive, be subjected to the influence of target shadow little, but can't detect the lap of object on suddenly static object and the consecutive frame.The background subtraction point-score can obtain more accurate target image, and operation efficiency is also than higher, and shortcoming is the scene that illumination and external condition cause to be changed relatively more responsive, can't take into account the turnover rate of background and the real-time of system.
Find through literature search prior art, Yu Jing etc. propose the chromatic value ordering of elder generation with each frame co-located pixel in the article " based on the moving object detection of chromaticity distortion " that " computer engineering " delivered on the monthly magazine in 2006 3, get median again and obtain background frames, and directly luminance difference and colour difference matrix are carried out the method for moving object detection by getting threshold value.This method has following shortcoming: carrying out can staying a large amount of foreground moving tracks when background frames extracts; Do not have filtering noise after obtaining luminance difference and colour difference matrix, this can influence the detection of follow-up foreground object, directly removes noise according to connected region, may also get rid of less moving object; Luminance difference and colour difference threshold value determination method are not clearly proposed; Judge that moving object has certain deviation because luminance difference and colour difference combine, this method does not clearly propose to keep object boundary.
Summary of the invention
The present invention is directed to the deficiency that prior art exists, propose a kind of video moving object detection method in conjunction with chromaticity distortion and luminance deviation.That the moving target detecting method that the present invention proposes can solve is few at the video frame number, foreground moving is slowly under the situation, the problem of background extracting poor effect; By the medium filtering of luminance difference and colour difference matrix being removed partial noise point and being kept the border; Propose a kind of method of definite luminance difference threshold value, can determine corresponding luminance difference threshold value accurately according to different video sequence characteristics; Keep the border of moving target by the projection of point set, make the moving object detection result more accurate.
The present invention is achieved by the following technical solutions, and concrete steps are as follows:
The first step, the extraction of background frames
In existing background frames extracting method, for R, the G of a certain pixel in the video sequence, the chromatic value of B passage, only when passing through this, the foreground moving object just big variation can take place.With the chromatic value ordering of each frame co-located pixel, get median and just can obtain the chromatic value of background frames in this point.But it is less to run into video sequence frame, and perhaps the slow situation of foreground object movement velocity has the vestige that moving target stays on the background frames that extracts.
In the present invention, the method for using statistics to add filtering is carried out the extraction of background frames, and concrete steps are as follows:
1. the method subchannel with statistics extracts background frames
The chromatic value scope of each passage of RGB all is [0,255], and this interval is divided into n part, is designated as [0, p], [p+1,2p] ..., [(n-1) p, 255].If chromatic value pixel (i, j, d, k) concentrated the be distributed in interval [mp+1 of certain pixel in each frame, (m+1) p] (0＜m＜n), then to being distributed in all pixel (i, the j in this interval, d, the background pixel chromatic value of median as the d passage got in k) value ordering.
2. medium filtering
At the pixel (i of subchannel to black track zone, j, d, when k) carrying out distribution statistics, ordering and selection, the chromatic value of one of them passage is mutually far short of what is expected with the chromatic value of two other passage, will demonstrate irregular color spot, color lump this moment: less such as green and blue chromatic value, and the red channel chromatic value is bigger, then the synthetic back of triple channel redness has just shown.By analysis, find that these noises meet the model of salt-pepper noise, so can adopt medium filtering to eliminate the interference of these points to the uneven point of these colors.
3. revise chromatic value
For eliminating the color spot of the fritter that exists in the background frames that 2. obtains by step, the former RGB triple channel chromatic value of saturation drastic change point is replaced to the three-channel colourity average of RGB.
Second step is in conjunction with the video frequency motion target detection of luminance difference and colour difference
The method that the present invention proposes, before utilizing luminance difference and colour difference information to detect foreground object, earlier luminance difference and colour difference matrix are carried out filtering, to eliminate a part of noise, utilize the Gauss model of statistics to calculate the threshold value of luminance difference again, and rule of thumb value is determined the threshold value of colour difference; At last the result who detects is made to keep BORDER PROCESSING, make testing result more accurate.Method step is as follows:
1. calculate luminance deviation and chromaticity distortion
If E
i=[E
R(i), E
G(i), E
B(i)] be the rgb value of i pixel of background image, I
i=I
R(i), I
G(i), I
B(i)] rgb value of expression present image i pixel, with the luminance deviation of following formula calculating present image and background image:
Calculate the chromaticity distortion of present image and background image with following formula:
2. the elimination of noise
Before carrying out detection method, luminance difference and the colour difference matrix to each frame carries out medium filtering earlier, removes a part of noise and keeps the border, does like this and can reduce the influence of noise spot to moving object detection, increases accuracy.
3. introduce probabilistic statistical method and determine the luminance difference threshold value
If (d is that (i j) in the luminance difference of d passage, can be normalized to standardized normal distribution with all luminance difference according to following formula to pixel in the k frame k) to pixel for i, j.
z＝(pixel(i，j，d，k)-μ)/σ
If get Z=3, and Φ (z＞Z)=0.13%, then the luminance difference probability of occurrence is moving object less than 0.13% point, determine threshold value thus: secretly the prospect threshold value in background is T
-α=-3 σ+μ; Bright prospect threshold value in background is T
α=3 σ+μ.
If foreground object is darker than background, judge that then the condition of moving target is as follows:
pixel(i，j，d，k)＜T
-α
If foreground object is brighter than background, judge that then the condition of moving target is as follows:
pixel(i，j，d，k)＞T
α
Use two kinds of threshold values, can detect the brightness value object darker and brighter on the width of cloth picture simultaneously, and need not use boundary segmentation earlier, each moving object is extracted separately carry out the threshold value analysis than background than background.
In addition, after process filtering of colour difference matrix and the normalization, also need a fixed threshold value to determine the object of motion.The method that the present invention proposes is by the different video sequence of research, and the empirical value that obtains the colour difference threshold value is 0.1 (the colour difference span after the normalization is [0,1]).
4. moving Object Segmentation
If the moving target that is 2. obtained by step is point set F
a, the prospect point set that is obtained by the difference of original frame of video and background frames is F
b, with F
aAnd F
bRelatively:
If ● F
aCertain segment boundary at F
bIn, then get F
aThe border as the border of moving target, and at F
aBeyond this segment boundary, the F that is split
bSubclass be exactly the shadow region of moving target;
If ● F
bCertain segment boundary at F
aIn, then get F
bBe the border of moving target, the F that is split
aSubclass give up as noise spot.
The prior art that effect of the present invention and Yu Jing etc. propose in " based on the moving object detection of chromaticity distortion " that " computer engineering " delivered on the monthly magazine in 2006 3 is compared: can be few at the video frame number, accurately extract the background frames that does not contain foreground information under the situation that foreground moving is slower, the number of pixel that can the movement locus of prospect is shared reduces more than 80%, has improved the accuracy that background frames extracts greatly; By the medium filtering of luminance difference and colour difference matrix being removed partial noise point and being kept the border; Proposed a kind of method of definite luminance difference threshold value, can determine corresponding luminance difference threshold value exactly according to different video sequence characteristics, and the border of having adopted the projecting method between the point set to preserve moving target; Can make the accuracy rate of moving object detection improve more than 30%.
Description of drawings
Fig. 1 is the inventive method flow chart
Fig. 2 is the flow chart that background frames extracts in the inventive method
Fig. 3 is the flow chart that the luminance difference threshold value is calculated in the inventive method
Embodiment
Below in conjunction with accompanying drawing embodiments of the invention are elaborated: present embodiment is being to implement under the prerequisite with the technical solution of the present invention, provided detailed execution mode and concrete operating process, but protection scope of the present invention is not limited to following embodiment.
As shown in Figure 1, present embodiment can be divided into following step.
(1) utilize statistical method to divide the RGB passage to extract background frames
In the present embodiment, the method for utilizing statistics to add filtering divides the RGB passage to extract background frames, its flow process as shown in Figure 2, concrete steps are as follows:
1. utilize statistical method to divide the RGB passage to extract background frames
The chromatic value scope of each passage of RGB all is [0,255], and this interval is divided into n part, is designated as [0, p], [p+1,2p] ..., [(n-1) p, 255].For the video sequence that will carry out motion detection, if the chromatic value pixel (i of certain pixel in each frame, j, d, k) the concentrated interval [mp+1 that is distributed in, (m+1) p] (0＜m＜n), then to being distributed in all pixel (i, j, the d in this interval, k) the background pixel chromatic value of median as the d passage got in value ordering.
2. medium filtering
Analyze result 1., can find to exist in the background frames the uneven point of a lot of colors, find by analysis, the uneven point of these colors meets the model of salt-pepper noise, so available medium filtering is eliminated the interference of the uneven point of color.
3. revise chromatic value
For object that still has foreground moving in the result 2. and the more unusual colored fritter of color, can eliminate by following processing: if the chromatic value of certain passage is obviously big or little than other passage, and and all around pixel colourity difference bigger, saturation changes greater than certain threshold value, then delete the original RGB passage chromatic value of this pixel, and replace with the colourity average of view picture background frames.
When the method that use Yu Jing etc. proposes in " based on the moving object detection of chromaticity distortion " that " computer engineering " delivered on the monthly magazine in 2006 3 is carried out the extraction of background frames, contain a large amount of foreground moving trace informations in the background frames, this means that background frames extracts not accurate enoughly, will directly have influence on the result of follow-up moving object detection.And the method for using present embodiment to propose can reduce the movement locus of prospect in a large number, under the condition especially limited at the video frame number, that foreground moving is slower, still can reach very desirable effect.
(2) luminance difference and the colourity difference of calculating present frame and background frames
If E
i=[E
R(i), E
G(i), E
B(i)] be the rgb value of i pixel of background image, I
i=I
R(i), I
G(i), I
B(i)] rgb value of expression present image i pixel, with the luminance deviation of following formula calculating present image and background image:
Calculate the chromaticity distortion of present image and background image with following formula:
(3) medium filtering
In the present embodiment, before utilizing luminance difference and colour difference information to detect foreground object, luminance difference and the colour difference matrix to each frame carries out medium filtering earlier, removes a part of noise and keeps the border.Do like this and can reduce the influence of noise spot, increase the accuracy that detects moving object detection.
(4) determine luminance difference and colour difference threshold value with probabilistic method
In the present invention, the calculation process that obtains the luminance difference threshold value with probabilistic method as shown in Figure 3.
1. normalization
Because image shows and the needs of subsequent treatment, all gray values must be normalized to [0,1] interval, in the luminance difference matrix of all frames of video sequence, the one part of pixel point of difference maximum is very big for the probability of noise, therefore get 90% of maximum difference, be designated as t, as the new maximum of all pixel intensity differences, and luminance difference is greater than the pixel of t, be designated as p, regard noise spot as, in the normalized process former luminance difference p is changed into new value t.Pass through normalized, can remove the noise of a part again.
2. use Gauss model match luminance difference matrix
The existence of noise has a strong impact on the accuracy of the luminance difference variance that each frame calculates, and therefore, must carry out medium filtering earlier.Then, average μ and variances sigma are asked in all pixel unifications of all frames.
If (d is that (i j) in the luminance difference of d passage, can be normalized to standardized normal distribution with all luminance difference according to following formula to pixel in the k frame k) to pixel for i, j.
z＝(pixel(i，j，d，k)-μ)/σ
Can determine thus:
Secretly the prospect threshold value in background is
T
-α＝-Z□σ+μ
Bright prospect threshold value in background is
T
α＝Z□σ+μ
For example, get and decide Z=3, (z＞Z)=0.13% thinks that the luminance difference probability of occurrence is moving object less than 0.13% point because Φ; So when z＜-judge that then this point is the foreground point when Z or z＞Z.
So just all luminance difference are normalized to standardized normal distribution, so and threshold value is decided by getting fixed Z, can calculate different luminance difference according to the characteristics of different video sequences, and need not be at every turn all determine by artificial.
3. calculated threshold
Foreground object is darker or bright than background, determines the luminance difference threshold value respectively by two kinds of situations:
● if foreground object is darker than background, judges that then the condition of moving target is as follows:
pixel(i，j，d，k)＜T
-α
● if foreground object is brighter than background, judges that then the condition of moving target is as follows:
pixel(i，j，d，k)＞T
α
In addition, after process filtering of colour difference matrix and the normalization, also need a fixed threshold value to determine the object of motion.The method that present embodiment proposes is by the different video sequence of research, and the empirical value that obtains the colour difference threshold value is 0.1 (the colour difference span after the normalization is [0,1]).
(5) detect the foreground moving object
Use two kinds of luminance difference threshold values and colour difference threshold value in (4), can detect the brightness value object darker and brighter on the width of cloth picture simultaneously, and need not use boundary segmentation earlier, each moving object is extracted separately carry out the threshold value analysis than background than background.
(6) testing result is kept BORDER PROCESSING
The edge of the moving target that detects for reservation need keep BORDER PROCESSING to testing result.If the moving target that is obtained by step (5) is point set F
a, the prospect point set that is obtained by the difference of original frame of video and background frames is F
bWith F
aAnd F
bRelatively:
If ● F
aCertain segment boundary at F
bIn, then get F
aThe border as the border of moving target, and at F
aBeyond this segment boundary, the F that is split
bSubclass be exactly the shadow region of moving target;
If ● F
bCertain segment boundary at F
aIn, then get F
bBe the border of moving target, the F that is split
aSubclass give up as noise spot.
Claims (5)
1, a kind of video moving object detection method in conjunction with chromaticity distortion and luminance deviation is characterized in that concrete steps are as follows:
The first step, the extraction of background frames
The method of using statistics to add filtering is carried out the extraction of background frames, at first the method subchannel with statistics extracts background frames, adopt medium filtering to eliminate the interference of the uneven point of color in the background frames of subchannel extraction then, former RGB triple channel chromatic value with saturation drastic change point replaces to the three-channel colourity average of RGB at last, to eliminate the color spot of the fritter that exists in the background frames that obtains behind the medium filtering;
Second step is in conjunction with the video frequency motion target detection of luminance difference and colour difference
Before utilizing luminance difference and colour difference information to detect foreground object, earlier luminance difference and colour difference matrix are carried out filtering, to eliminate a part of noise, utilize the Gauss model of adding up to calculate the threshold value of luminance difference again, and rule of thumb value is determined the threshold value of colour difference, at last the result who detects is made to keep BORDER PROCESSING.
2, video moving object detection method in conjunction with chromaticity distortion and luminance deviation according to claim 1 is characterized in that, described method subchannel with statistics extracts background frames, be specially: the chromatic value scope of each passage of RGB all is [0,255], and this interval is divided into n part, be designated as [0, p], [p+1,2p] ..., [(n-1) p, 255], if chromatic value pixel (i, j, the d of certain pixel in each frame, k) the concentrated interval [mp+1 that is distributed in, (m+1) p], 0＜m＜n is then to being distributed in all the pixel (i in this interval, j, d, the background pixel chromatic value of median as the d passage got in k) value ordering.
3, the video moving object detection method in conjunction with chromaticity distortion and luminance deviation according to claim 1 is characterized in that, described video frequency motion target in conjunction with luminance difference and colour difference detects, and step is as follows:
1. calculate luminance deviation and chromaticity distortion
If E
i=[E
R(i), E
G(i), E
B(i)] be the rgb value of i pixel of background image, I
i=[I
R(i), I
G(i), I
B(i)] rgb value of expression present image i pixel, with the luminance deviation of following formula calculating present image and background image:
Calculate the chromaticity distortion of present image and background image with following formula:
2. the elimination of noise
Luminance difference and colour difference matrix to each frame carries out medium filtering earlier, removes a part of noise and keeps the border;
3. introduce probabilistic statistical method and determine the luminance difference threshold value
If (d is that (i j) in the luminance difference of d passage, is normalized to standardized normal distribution according to following formula with all luminance difference to pixel in the k frame k) to pixel for i, j;
z＝(pixel(i，j，d，k)-μ)/σ
If get Z=3, and Φ (z＞Z)=0.13%, then the luminance difference probability of occurrence is moving object less than 0.13% point, determine threshold value thus: secretly the prospect threshold value in background is T
-α=-3 σ+μ; Bright prospect threshold value in background is T
α=3 σ+μ;
If foreground object is darker than background, judge that then the condition of moving target is as follows:
pixel(i，j，d，k)＜T
-α
If foreground object is brighter than background, judge that then the condition of moving target is as follows:
pixel(i，j，d，k)＞T
α
4. moving Object Segmentation
If the moving target that is 2. obtained by step is point set F
a, the prospect point set that is obtained by the difference of original frame of video and background frames is F
b, with F
aAnd F
bRelatively to determine the border of moving target.
4, the video moving object detection method in conjunction with chromaticity distortion and luminance deviation according to claim 3, it is characterized in that, after process filtering of colour difference matrix and the normalization, set a threshold value and determine the object of motion, the empirical value of colour difference threshold value is 0.1, colour difference span after the normalization is [0,1].
5, the video moving object detection method in conjunction with chromaticity distortion and luminance deviation according to claim 3 is characterized in that, and is described with F
aAnd F
bRelatively, be specially to determine the border of moving target:
If F
aCertain segment boundary at F
bIn, then get F
aThe border as the border of moving target, and at F
aBeyond this segment boundary, the F that is split
bSubclass be exactly the shadow region of moving target;
If F
bCertain segment boundary at F
aIn, then get F
bBe the border of moving target, the F that is split
aSubclass give up as noise spot.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007100437348A CN100531374C (en) | 2007-07-12 | 2007-07-12 | Chroma deviation and brightness deviation combined video moving object detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007100437348A CN100531374C (en) | 2007-07-12 | 2007-07-12 | Chroma deviation and brightness deviation combined video moving object detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101098462A true CN101098462A (en) | 2008-01-02 |
| CN100531374C CN100531374C (en) | 2009-08-19 |
Family
ID=39011928
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2007100437348A Expired - Fee Related CN100531374C (en) | 2007-07-12 | 2007-07-12 | Chroma deviation and brightness deviation combined video moving object detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100531374C (en) |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101859440A (en) * | 2010-05-31 | 2010-10-13 | 浙江捷尚视觉科技有限公司 | Block-based motion region detection method |
| CN102184553A (en) * | 2011-05-24 | 2011-09-14 | 杭州华三通信技术有限公司 | Moving shadow detecting method and device |
| CN102193651A (en) * | 2010-03-11 | 2011-09-21 | 株式会社理光 | Equipment, method and system for identifying laser point |
| CN102446346A (en) * | 2010-09-30 | 2012-05-09 | 北京中电兴发科技有限公司 | Method for quickly removing motion image shadow |
| CN102542572A (en) * | 2010-12-20 | 2012-07-04 | 财团法人工业技术研究院 | Method for detecting moving object image in video sequence and image processing system |
| CN102609902A (en) * | 2010-12-10 | 2012-07-25 | 卡西欧计算机株式会社 | Image processing apparatus, image processing method, and storage medium |
| CN103116889A (en) * | 2013-02-05 | 2013-05-22 | 海信集团有限公司 | Positioning method and electronic device |
| CN102054270B (en) * | 2009-11-10 | 2013-06-05 | 华为技术有限公司 | Method and device for extracting foreground from video image |
| WO2015014286A1 (en) * | 2013-07-31 | 2015-02-05 | 华为终端有限公司 | Method and apparatus for generating high dynamic range image |
| CN104349125A (en) * | 2013-08-05 | 2015-02-11 | 浙江大华技术股份有限公司 | Area monitoring method and device |
| CN104751487A (en) * | 2015-03-26 | 2015-07-01 | 太原理工大学 | Method for detecting movement target based on colored RGB three-pane color-change frame difference |
| CN105100586A (en) * | 2014-05-21 | 2015-11-25 | 卡西欧计算机株式会社 | Detection device and detection method |
| CN105183755A (en) * | 2015-07-17 | 2015-12-23 | 小米科技有限责任公司 | Picture display method and apparatus |
| CN105261021A (en) * | 2015-10-19 | 2016-01-20 | 浙江宇视科技有限公司 | Method and apparatus of removing foreground detection result shadows |
| CN106127214A (en) * | 2016-06-30 | 2016-11-16 | 四川大学 | A kind of monitor video robust background modeling method based on linear projection and device |
| CN106651824A (en) * | 2015-10-28 | 2017-05-10 | 富士通株式会社 | Shadow detection device and shadow detection method |
| CN107222962A (en) * | 2017-07-06 | 2017-09-29 | 东莞市商宝照明技术有限公司 | Illumination control apparatus, method and system |
| CN109086761A (en) * | 2018-06-28 | 2018-12-25 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium, electronic equipment |
| CN109544605A (en) * | 2018-05-23 | 2019-03-29 | 安徽大学 | Mobile shadow detection method based on time-space relationship modeling |
| CN109903556A (en) * | 2019-03-01 | 2019-06-18 | 成都众易通科技有限公司 | A kind of vehicle blind zone on-line monitoring early warning system |
| CN109998364A (en) * | 2019-05-16 | 2019-07-12 | 珠海格力电器股份有限公司 | Classification cooking control method, control equipment, cooking equipment, terminal and system |
| CN111582268A (en) * | 2020-04-13 | 2020-08-25 | 浙江大华技术股份有限公司 | License plate image processing method and device and computer storage medium |
| CN112950484A (en) * | 2019-12-11 | 2021-06-11 | 鸣医(上海)生物科技有限公司 | Method for removing color pollution of photographic image |
| CN116977228A (en) * | 2023-09-25 | 2023-10-31 | 广东匠芯创科技有限公司 | Image noise reduction method, electronic device and storage medium |
| CN117629514A (en) * | 2024-01-26 | 2024-03-01 | 吉林大学 | SF6 leakage amount detection system and method based on mid-infrared thermal imaging |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5201203B2 (en) * | 2010-12-22 | 2013-06-05 | カシオ計算機株式会社 | Image processing apparatus, image processing method, and program |
-
2007
- 2007-07-12 CN CNB2007100437348A patent/CN100531374C/en not_active Expired - Fee Related
Cited By (46)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102054270B (en) * | 2009-11-10 | 2013-06-05 | 华为技术有限公司 | Method and device for extracting foreground from video image |
| CN102193651A (en) * | 2010-03-11 | 2011-09-21 | 株式会社理光 | Equipment, method and system for identifying laser point |
| CN102193651B (en) * | 2010-03-11 | 2014-11-05 | 株式会社理光 | Equipment, method and system for identifying laser point |
| CN101859440A (en) * | 2010-05-31 | 2010-10-13 | 浙江捷尚视觉科技有限公司 | Block-based motion region detection method |
| CN102446346A (en) * | 2010-09-30 | 2012-05-09 | 北京中电兴发科技有限公司 | Method for quickly removing motion image shadow |
| CN102609902B (en) * | 2010-12-10 | 2015-04-08 | 卡西欧计算机株式会社 | Image processing apparatus, image processing method |
| CN102609902A (en) * | 2010-12-10 | 2012-07-25 | 卡西欧计算机株式会社 | Image processing apparatus, image processing method, and storage medium |
| CN102542572B (en) * | 2010-12-20 | 2014-09-17 | 财团法人工业技术研究院 | Method for detecting moving object image in video sequence and image processing system |
| CN102542572A (en) * | 2010-12-20 | 2012-07-04 | 财团法人工业技术研究院 | Method for detecting moving object image in video sequence and image processing system |
| US8922651B2 (en) | 2010-12-20 | 2014-12-30 | Industrial Technology Research Institute | Moving object detection method and image processing system for moving object detection |
| CN102184553A (en) * | 2011-05-24 | 2011-09-14 | 杭州华三通信技术有限公司 | Moving shadow detecting method and device |
| CN103116889A (en) * | 2013-02-05 | 2013-05-22 | 海信集团有限公司 | Positioning method and electronic device |
| WO2015014286A1 (en) * | 2013-07-31 | 2015-02-05 | 华为终端有限公司 | Method and apparatus for generating high dynamic range image |
| CN104349066A (en) * | 2013-07-31 | 2015-02-11 | 华为终端有限公司 | Method and device for generating images with high dynamic ranges |
| CN104349066B (en) * | 2013-07-31 | 2018-03-06 | 华为终端（东莞）有限公司 | A kind of method, apparatus for generating high dynamic range images |
| CN104349125A (en) * | 2013-08-05 | 2015-02-11 | 浙江大华技术股份有限公司 | Area monitoring method and device |
| CN104349125B (en) * | 2013-08-05 | 2017-10-27 | 浙江大华技术股份有限公司 | A kind of area monitoring method and equipment |
| CN105100586B (en) * | 2014-05-21 | 2018-11-16 | 卡西欧计算机株式会社 | Detection device, detection method and recording medium |
| CN105100586A (en) * | 2014-05-21 | 2015-11-25 | 卡西欧计算机株式会社 | Detection device and detection method |
| CN104751487B (en) * | 2015-03-26 | 2017-06-06 | 太原理工大学 | A kind of moving target detecting method based on the plane discoloration frames of colored RGB tri- difference |
| CN104751487A (en) * | 2015-03-26 | 2015-07-01 | 太原理工大学 | Method for detecting movement target based on colored RGB three-pane color-change frame difference |
| CN105183755A (en) * | 2015-07-17 | 2015-12-23 | 小米科技有限责任公司 | Picture display method and apparatus |
| WO2017012232A1 (en) * | 2015-07-17 | 2017-01-26 | 小米科技有限责任公司 | Method and apparatus for displaying picture |
| CN105183755B (en) * | 2015-07-17 | 2018-10-12 | 小米科技有限责任公司 | A kind of display methods and device of picture |
| RU2647673C2 (en) * | 2015-07-17 | 2018-03-16 | Сяоми Инк. | Method and apparatus for displaying picture |
| US9892524B2 (en) | 2015-07-17 | 2018-02-13 | Xiaomi Inc. | Image displaying methods, devices, and storage media |
| CN105261021B (en) * | 2015-10-19 | 2019-03-08 | 浙江宇视科技有限公司 | Remove the method and device of foreground detection result shade |
| CN105261021A (en) * | 2015-10-19 | 2016-01-20 | 浙江宇视科技有限公司 | Method and apparatus of removing foreground detection result shadows |
| CN106651824A (en) * | 2015-10-28 | 2017-05-10 | 富士通株式会社 | Shadow detection device and shadow detection method |
| CN106127214B (en) * | 2016-06-30 | 2018-07-17 | 四川大学 | A kind of monitor video robust background modeling method and device based on linear projection |
| CN106127214A (en) * | 2016-06-30 | 2016-11-16 | 四川大学 | A kind of monitor video robust background modeling method based on linear projection and device |
| CN107222962A (en) * | 2017-07-06 | 2017-09-29 | 东莞市商宝照明技术有限公司 | Illumination control apparatus, method and system |
| CN107222962B (en) * | 2017-07-06 | 2023-06-16 | 东莞市商宝照明技术有限公司 | Lighting control device, method and system |
| CN109544605B (en) * | 2018-05-23 | 2022-01-21 | 安徽大学 | Moving shadow detection method based on space-time relation modeling |
| CN109544605A (en) * | 2018-05-23 | 2019-03-29 | 安徽大学 | Mobile shadow detection method based on time-space relationship modeling |
| CN109086761A (en) * | 2018-06-28 | 2018-12-25 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium, electronic equipment |
| CN109086761B (en) * | 2018-06-28 | 2020-12-01 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN109903556A (en) * | 2019-03-01 | 2019-06-18 | 成都众易通科技有限公司 | A kind of vehicle blind zone on-line monitoring early warning system |
| CN109998364A (en) * | 2019-05-16 | 2019-07-12 | 珠海格力电器股份有限公司 | Classification cooking control method, control equipment, cooking equipment, terminal and system |
| CN112950484A (en) * | 2019-12-11 | 2021-06-11 | 鸣医(上海)生物科技有限公司 | Method for removing color pollution of photographic image |
| CN112950484B (en) * | 2019-12-11 | 2023-06-16 | 鸣医(上海)生物科技有限公司 | Method for removing color pollution of photographic image |
| CN111582268A (en) * | 2020-04-13 | 2020-08-25 | 浙江大华技术股份有限公司 | License plate image processing method and device and computer storage medium |
| CN111582268B (en) * | 2020-04-13 | 2023-05-12 | 浙江大华技术股份有限公司 | License plate image processing method and device and computer storage medium |
| CN116977228A (en) * | 2023-09-25 | 2023-10-31 | 广东匠芯创科技有限公司 | Image noise reduction method, electronic device and storage medium |
| CN116977228B (en) * | 2023-09-25 | 2024-02-09 | 广东匠芯创科技有限公司 | Image noise reduction method, electronic device and storage medium |
| CN117629514A (en) * | 2024-01-26 | 2024-03-01 | 吉林大学 | SF6 leakage amount detection system and method based on mid-infrared thermal imaging |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100531374C (en) | 2009-08-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100531374C (en) | Chroma deviation and brightness deviation combined video moving object detection method | |
| CN101084527B (en) | A method and system for processing video data | |
| CN101236606B (en) | Shadow cancelling method and system in vision frequency monitoring | |
| US10445567B2 (en) | Pedestrian head identification method and system | |
| CN103714325B (en) | Left object and lost object real-time detection method based on embedded system | |
| Jung | Efficient background subtraction and shadow removal for monochromatic video sequences | |
| TWI409718B (en) | Method of locating license plate of moving vehicle | |
| CN101303727B (en) | Intelligent management method based on video human number Stat. and system thereof | |
| US7639878B2 (en) | Shadow detection in images | |
| CN102567727B (en) | Method and device for replacing background target | |
| CN105005766B (en) | A kind of body color recognition methods | |
| CN112036254B (en) | Moving vehicle foreground detection method based on video image | |
| US20070036432A1 (en) | Object detection in images | |
| CN103258232B (en) | A kind of public place crowd estimate's method based on dual camera | |
| CN105404847A (en) | Real-time detection method for object left behind | |
| CN109218695A (en) | Video image enhancing method, device, analysis system and storage medium | |
| CN101364304B (en) | Shadow detection method based on color invariance and Gauss model | |
| CN104951742A (en) | Detection method and system for sensitive video | |
| Tiwari et al. | Shadow detection and compensation in aerial images using MATLAB | |
| CN102610104B (en) | Onboard front vehicle detection method | |
| Chen et al. | Robust license plate detection in nighttime scenes using multiple intensity IR-illuminator | |
| CN112446913B (en) | Method, device and equipment for counting pedestrian number and judging trailing of passway gate | |
| CN106844394A (en) | Based on the video retrieval method that the upper and lower clothing color of pedestrian differentiates | |
| JP4697923B2 (en) | Counting system and counting method for moving object in water or water surface | |
| CN110688979A (en) | Illegal vehicle tracking method and device |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090819 Termination date: 20130712 |