CN103729828A - Video rain removing method - Google Patents
Video rain removing method Download PDFInfo
- Publication number
- CN103729828A CN103729828A CN201310683538.2A CN201310683538A CN103729828A CN 103729828 A CN103729828 A CN 103729828A CN 201310683538 A CN201310683538 A CN 201310683538A CN 103729828 A CN103729828 A CN 103729828A
- Authority
- CN
- China
- Prior art keywords
- video
- value
- illumination variation
- calculate
- frame
- 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
Images
Abstract
The invention provides a video rain removing method. According to the video rain removing method, illumination variation factors are taken into consideration, firstly, whether environment illumination varies or not is judged, on the basis, rain removing under the illumination variation condition is achieved according to the illumination variation by changing the Kalman gain Kg, whether pixels under illumination are covered by raindrops or not is judged according to the difference of the brightness measurement value of a current frame and the brightness actual value of a previous frame, and therefore the raindrops are removed. According to the video rain removing method, the environment illumination variation factors are taken into consideration in video rain removing, and video rain removing can be effectively performed when illumination variation exists in the environment.
Description
Technical field
The present invention relates to technical field of computer vision, relate in particular to a kind of video raindrop removal method that contains illumination variation factor.
Background technology
To video image, imaging has a great impact rain, can cause the covering of the fuzzy and information of video image imaging, and its direct result is that the sharpness of video image declines, and the digitized processing of video image is affected and hydraulic performance decline by this also can.The video image that polluted by raindrop is carried out to the further processing that repair process is conducive to video image.And the target detection of video image, tracking, identification or cutting techniques are all used widely in a plurality of fields such as modern military, traffic and security monitorings.
Video goes rain technology to propose to have obtained till now significant progress from 2003, and the various methods based on different mathematics physics model are successively proposed by scholars, and the effect that raindrop are removed is also enhanced gradually.Existing going in rain method, algorithm based on video successive frame has incorporated the temporal information of video, effect when it removes video raindrop be generally also better than based on single-frame images go rain method, particularly more remarkable on to the retention at background object edge.The go rain method of picture based on the pixel intensity degree of bias (skew), gray tone, K mean cluster, Kalman filter, fuzzy connectedness is all more novel, effective.The Kalman filter of take is below introduced a kind of method of carrying out raindrop removal based on multi-frame video image as example.
Kalman filter is a kind of optimization autoregression data processing algorithm.For following formula, represent discrete control procedure system:
X(k)=AX(k-1)+BU(k)+W(k) （1）
Systematic survey value:
Z(k)=HX(k)+V(k) （2）
Wherein, system state when X (k), X (k-1) represent respectively time k and k-1, U (k) expression system is controlled parameter, and W (k) and V (k) represent respectively the noise of process and measurement, and they meet Gaussian distribution.A, B are systematic parameter, and H is measuring system parameter, and for multifactor system, they are all matrixes.
For meeting the system of two linear random differential equations above, can carry out optimal estimation by Kalman filter.The core of Kalman filter is following 5 equatioies:
X(k|k-1)=AX(k-1|k-1)+BU(k) （3）
P(k|k-1)=AP(k-1|k-1)A'+Q （4）
X(k|k)=X(k|k-1)+Kg(k)(Z(k)-HX(k|k-1)) （5）
Kg(k)=P(k|k-1)H'/(HP(k|k-1)H'+R) （6）
P(k|k)=(I-Kg(k)H)P(k|k-1) （7）
Wherein, the variance of the corresponding X of P (k|k) (k|k), Q is the variance of systematic procedure, Kg (k) is called kalman gain (Kalman Gain).Kalman filter in simple terms, is estimated actual value by system prediction value and measured value exactly.
When Kalman wave filter is used for removing raindrop, establish initial value X (0)=100, A=1, H (0)=1, P (0)=1.Because raindrop can be regarded the noise that variance is very large as, therefore the initial variance of V (k) is made as to 50, the initial variance of W (k) is set to 5 simultaneously.Side's extent has determined to reach speed and the removal effect quality to noise of estimated value, and variance is little means that to reach predicted value slower, but removal effect is better.
In prior art, the rain method of going based on video continuous multiple frames monochrome information is not all considered the factor that ambient lighting changes, if therefore there is illumination variation in practice, will have influence on to a great extent rain effect.Take Kalman filter as example, if a certain moment k intensity of illumination increases, " actual value " calculating according to above-mentioned formula (5) pixel intensity like this will be little more a lot of than real actual value.
Summary of the invention
For the problems referred to above, the object of this invention is to provide a kind of video that contains illumination variation factor and go rain method, to increase existing video, go the robustness of rain method.
Video goes a rain method, it is characterized in that, comprises the steps:
S11, calculate the poor of the brightness measurement value of present frame and the brightness actual value of former frame, as the first element and build the first matrix;
S12, calculate the average of other the first element sums within the scope of surrounding's predeterminable area of the first element described in each, as the second element and build the second matrix;
, if the absolute value of described the second element is greater than the threshold values of described illumination variation, there is illumination variation in the absolute value of S13, more described the second element and the threshold values of illumination variation, carries out S14, otherwise carry out S16;
S14, the element calculating between described the first element and described the second element are poor, and compare with pre-set threshold value, if described element is poor, be greater than described pre-set threshold value, judge that pixel corresponding to described the first element covered by raindrop, and carry out S16, otherwise carry out S15;
S15, utilize Kalman filter to calculate the intrinsic brilliance value of described present frame, and change kalman gain into Kg=Kg+a*|y|+b, wherein a, b are for being used for the normalized parameter of described the second element;
S16, utilize Kalman filter to calculate the intrinsic brilliance value of described present frame;
S17, execution next frame are until finish.
In the present invention's one better embodiment, before step S11, also comprise: S10, read in video, obtain the brightness actual value of described former frame.
In the present invention's one better embodiment, before step S11, also comprise: the initial value that Kalman filter is set.
In the present invention's one better embodiment, the square area that described predeterminable area scope is 21x21.
Compared to prior art, video provided by the invention goes rain method to remove to have considered in the rain ambient lighting changing factor at video, can when environment exists illumination variation, effectively carry out video and remove rain.
Above-mentioned explanation is only the general introduction of technical solution of the present invention, in order to better understand technological means of the present invention, and can be implemented according to the content of instructions, and for above and other objects of the present invention, feature and advantage can be become apparent, below especially exemplified by embodiment, and coordinate accompanying drawing, be described in detail as follows.
Accompanying drawing explanation
The video that Fig. 1 provides for a preferred embodiment of the present invention removes the process flow diagram of rain method.
Embodiment
Below in conjunction with drawings and the specific embodiments, the present invention is further detailed explanation.
Refer to Fig. 1, one embodiment of the invention provides a kind of video to go rain method, and it comprises the steps:
S11, calculate the poor of the brightness measurement value of present frame and the brightness actual value of former frame, as the first element and build the first matrix.
Particularly, obtain the poor of the brightness measurement value of present frame and brightness actual value that former frame calculates, before and after the luminance difference of two frames, as the first element, deposit the first matrix x in.Be understandable that, described the first matrix x includes a plurality of the first elements.
In the present embodiment, before step S11, also comprise: S10, read in video, and obtain the brightness actual value of described former frame.
Also comprise step (not shown): the initial value that Kalman filter is set thereafter.Establish initial value X (0)=100, A=1, H (0)=1, P (0)=1, concrete can, with reference to the content of background technology of the present invention, repeat no more herein.
S12, calculate the average of other the first element sums within the scope of surrounding's predeterminable area of the first element described in each, as the second element and build the second matrix.
In the present embodiment, the square area that described predeterminable area scope is 21x21.Particularly, calculate the average of other the first element sums in 21x21 square area around of each first element in described the first matrix x, as the second element, deposit the second matrix y in.Certainly, described predeterminable area scope is not limited in the present embodiment, also can set as required, as long as square area meets: on the one hand, square area is enough to be caused by a large amount of raindrop to get rid of that its brightness increases greatly; The regional area illumination variation that while on the other hand, avoiding square area too large, pointolite causes causes other region also to have the erroneous judgement of illumination variation.
Be understandable that, described the second matrix y also includes a plurality of the second elements.
, if the absolute value of described the second element is greater than the threshold values of described illumination variation, there is illumination variation in the absolute value of S13, more described the second element and the threshold values of illumination variation, carries out S14, otherwise carry out S16.
In the present embodiment, for each second element in described the second matrix y, judged whether | y|>c, there is the threshold value of illumination variation in c representative wherein, when | y| thinks and has illumination variation when very large, execution step S14, otherwise execution step S16.
S14, the element calculating between described the first element and described the second element are poor, and compare with pre-set threshold value, if described element is poor, be greater than described pre-set threshold value, judge that pixel corresponding to described the first element covered by raindrop, and carry out S16, otherwise carry out S15.
In the present embodiment, judged whether (x-y) >T, wherein, T is the threshold value that can artificially set.Thus, can judge whether x is covered by raindrop, if set up, illustrate that x is very large, now can judge that x is covered by raindrop, execution step S16, otherwise execution step S15.
S15, utilize Kalman filter to calculate the intrinsic brilliance value of described present frame, and change kalman gain into Kg=Kg+a*|y|+b, wherein a, b are for being used for the normalized parameter of described the second element.
In the present embodiment, the formula according to the present invention in background technology (3)～(7) are calculated the brightness actual value (repeating no more) of present frame herein, and change Kg into Kg=Kg+a*|y|+b, and wherein a, b are for by the normalized parameter of y.
Be understandable that, when | during y|>c, no matter represent illumination variation, be that illumination strengthens (y>0) or illumination weakens (y<0), all will increase the weight Kg of measured value Z (k) to adapt to illumination variation.
S16, utilize Kalman filter to calculate the intrinsic brilliance value of described present frame.
In the present embodiment, the formula according to the present invention in background technology (3)～(7) are calculated the brightness actual value (repeating no more) of present frame herein, and next frame is not ready.
S17, execution next frame are until finish.
Be understandable that, described video provided by the invention goes rain method to consider illumination variation factor, first judge whether ambient lighting changes, according to illumination variation, by changing kalman gain Kg realization, there is rainy the going of illumination variation situation on this basis, and by the extent of the brightness measurement value of present frame and the brightness actual value of former frame, judge under illumination, whether pixel is covered by raindrop, and then remove raindrop.
Compared to prior art, described video provided by the invention goes rain method to remove to have considered in the rain ambient lighting changing factor at video, can when environment exists illumination variation, effectively carry out video and remove rain.
The above, only embodiments of the invention, not the present invention is done to any pro forma restriction, although the present invention discloses as above with embodiment, yet not in order to limit the present invention, any those skilled in the art, do not departing within the scope of technical solution of the present invention, when can utilizing the technology contents of above-mentioned announcement to make a little change or being modified to the equivalent embodiment of equivalent variations, in every case be not depart from technical solution of the present invention content, any simple modification of above embodiment being done according to technical spirit of the present invention, equivalent variations and modification, all still belong in the scope of technical solution of the present invention.
Claims (4)
1. video goes a rain method, it is characterized in that, comprises the steps:
S11, calculate the poor of the brightness measurement value of present frame and the brightness actual value of former frame, as the first element and build the first matrix;
S12, calculate the average of other the first element sums within the scope of surrounding's predeterminable area of the first element described in each, as the second element and build the second matrix;
, if the absolute value of described the second element is greater than the threshold values of described illumination variation, there is illumination variation in the absolute value of S13, more described the second element and the threshold values of illumination variation, carries out S14, otherwise carry out S16;
S14, the element calculating between described the first element and described the second element are poor, and compare with pre-set threshold value, if described element is poor, be greater than described pre-set threshold value, judge that pixel corresponding to described the first element covered by raindrop, and carry out S16, otherwise carry out S15;
S15, utilize Kalman filter to calculate the intrinsic brilliance value of described present frame, and change kalman gain into Kg=Kg+a*|y|+b, wherein a, b are for being used for the normalized parameter of described the second element;
S16, utilize Kalman filter to calculate the intrinsic brilliance value of described present frame;
S17, execution next frame are until finish.
2. video as claimed in claim 1 goes rain method, it is characterized in that, before step S11, also comprises: S10, read in video, obtain the brightness actual value of described former frame.
3. video as claimed in claim 1 goes rain method, it is characterized in that, before step S11, also comprises: the initial value that Kalman filter is set.
4. video as claimed in claim 1 goes rain method, it is characterized in that, the square area that described predeterminable area scope is 21x21.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310683538.2A CN103729828B (en) | 2013-12-12 | 2013-12-12 | video rain removing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310683538.2A CN103729828B (en) | 2013-12-12 | 2013-12-12 | video rain removing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103729828A true CN103729828A (en) | 2014-04-16 |
| CN103729828B CN103729828B (en) | 2016-06-29 |
Family
ID=50453889
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310683538.2A Active CN103729828B (en) | 2013-12-12 | 2013-12-12 | video rain removing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103729828B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104299214A (en) * | 2014-09-30 | 2015-01-21 | 中国科学院深圳先进技术研究院 | Method and system for detecting and removing raindrops in light rain scene video data |
| CN104318537A (en) * | 2014-09-30 | 2015-01-28 | 中国科学院深圳先进技术研究院 | Method and system for detecting and removing raindrop in heavy rain scene video data |
| CN104978720A (en) * | 2015-07-01 | 2015-10-14 | 深圳先进技术研究院 | Video image raindrop removal method and apparatus |
| CN105184761A (en) * | 2015-08-28 | 2015-12-23 | 中国科学院深圳先进技术研究院 | Image rain removing method based on wavelet analysis and system |
| CN106023112A (en) * | 2016-05-24 | 2016-10-12 | 中国科学院深圳先进技术研究院 | Image rain removing method and system based on wavelet analysis |
| CN106067163A (en) * | 2016-05-24 | 2016-11-02 | 中国科学院深圳先进技术研究院 | A kind of image rain removing method based on wavelet analysis and system |
| CN109360155A (en) * | 2018-08-17 | 2019-02-19 | 上海交通大学 | Single-frame images rain removing method based on multi-scale feature fusion |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012066564A1 (en) * | 2010-11-15 | 2012-05-24 | Indian Institute Of Technology, Kharagpur | Method and apparatus for detection and removal of rain from videos using temporal and spatiotemporal properties. |
| US20130236116A1 (en) * | 2012-03-08 | 2013-09-12 | Industrial Technology Research Institute | Method and apparatus for single-image-based rain streak removal |
| CN103337061A (en) * | 2013-07-18 | 2013-10-02 | 厦门大学 | Rain and snow removing method for image based on multiple guided filtering |
-
2013
- 2013-12-12 CN CN201310683538.2A patent/CN103729828B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012066564A1 (en) * | 2010-11-15 | 2012-05-24 | Indian Institute Of Technology, Kharagpur | Method and apparatus for detection and removal of rain from videos using temporal and spatiotemporal properties. |
| US20130236116A1 (en) * | 2012-03-08 | 2013-09-12 | Industrial Technology Research Institute | Method and apparatus for single-image-based rain streak removal |
| CN103337061A (en) * | 2013-07-18 | 2013-10-02 | 厦门大学 | Rain and snow removing method for image based on multiple guided filtering |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104299214A (en) * | 2014-09-30 | 2015-01-21 | 中国科学院深圳先进技术研究院 | Method and system for detecting and removing raindrops in light rain scene video data |
| CN104318537A (en) * | 2014-09-30 | 2015-01-28 | 中国科学院深圳先进技术研究院 | Method and system for detecting and removing raindrop in heavy rain scene video data |
| CN104318537B (en) * | 2014-09-30 | 2017-07-28 | 中国科学院深圳先进技术研究院 | The detection of raindrop and minimizing technology and system in heavy rain scene video data |
| CN104299214B (en) * | 2014-09-30 | 2017-12-29 | 中国科学院深圳先进技术研究院 | The detection of raindrop and minimizing technology and system in light rain scene video data |
| CN104978720A (en) * | 2015-07-01 | 2015-10-14 | 深圳先进技术研究院 | Video image raindrop removal method and apparatus |
| CN105184761A (en) * | 2015-08-28 | 2015-12-23 | 中国科学院深圳先进技术研究院 | Image rain removing method based on wavelet analysis and system |
| CN106023112A (en) * | 2016-05-24 | 2016-10-12 | 中国科学院深圳先进技术研究院 | Image rain removing method and system based on wavelet analysis |
| CN106067163A (en) * | 2016-05-24 | 2016-11-02 | 中国科学院深圳先进技术研究院 | A kind of image rain removing method based on wavelet analysis and system |
| CN109360155A (en) * | 2018-08-17 | 2019-02-19 | 上海交通大学 | Single-frame images rain removing method based on multi-scale feature fusion |
| CN109360155B (en) * | 2018-08-17 | 2020-10-13 | 上海交通大学 | Single-frame image rain removing method based on multi-scale feature fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103729828B (en) | 2016-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103729828A (en) | Video rain removing method | |
| CN107301623B (en) | Traffic image defogging method and system based on dark channel and image segmentation | |
| CN103747213B (en) | A kind of real-time defogging method of the Traffic Surveillance Video based on moving target | |
| CN102768760B (en) | Quick image dehazing method on basis of image textures | |
| Gao et al. | Sand-dust image restoration based on reversing the blue channel prior | |
| EP3076367A1 (en) | Method for road detection from one image | |
| CN104616290A (en) | Target detection algorithm in combination of statistical matrix model and adaptive threshold | |
| CN103578083B (en) | Single image defogging method based on associating average drifting | |
| CN107301624A (en) | The convolutional neural networks defogging algorithm pre-processed based on region division and thick fog | |
| CN104599256A (en) | Single-image based image rain streak eliminating method and system | |
| CN104952256A (en) | Video information based method for detecting vehicles at intersection | |
| CN104915940A (en) | Alignment-based image denoising method and system | |
| CN109255799B (en) | Target tracking method and system based on spatial adaptive correlation filter | |
| CN104537622A (en) | Method and system for removing raindrop influence in single image | |
| CN111062331A (en) | Mosaic detection method and device for image, electronic equipment and storage medium | |
| CN110188640B (en) | Face recognition method, face recognition device, server and computer readable medium | |
| CN108898561B (en) | Defogging method, server and system for foggy image containing sky area | |
| CN104484865A (en) | Method for removing raindrops in video image | |
| CN109658441B (en) | Foreground detection method and device based on depth information | |
| CN103714518A (en) | Video rain removing method | |
| CN104537637A (en) | Method and device for estimating depth of single static image | |
| CN110580706A (en) | Method and device for extracting video background model | |
| CN111382736B (en) | License plate image acquisition method and device | |
| CN111079598B (en) | Lane line detection method based on image texture and machine learning | |
| TWI478078B (en) | A motion detection device and a motion detection method |
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 |