CN101510306A - Estimation method for video image illumination distribution - Google Patents
Estimation method for video image illumination distribution Download PDFInfo
- Publication number
- CN101510306A CN101510306A CNA2008101478621A CN200810147862A CN101510306A CN 101510306 A CN101510306 A CN 101510306A CN A2008101478621 A CNA2008101478621 A CN A2008101478621A CN 200810147862 A CN200810147862 A CN 200810147862A CN 101510306 A CN101510306 A CN 101510306A
- Authority
- CN
- China
- Prior art keywords
- image
- gradient
- illumination
- imgl
- gradient vector
- 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
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses an illumination distribution evaluation method of a video image, comprising the steps as follows: by gradient screening, the gradient vector influence on a part where a gradient magnitude is larger than a gradient threshold value or on edges is eliminated; and an illumination image reconstruction is carried out to a gradient vector image after screening so as to obtain an illumination image, thereby effectively eliminating the influence on the statistics of image illumination distribution, which is caused by the serious change of reflectivity of a scene surface, and accurately evaluating the characteristics and distribution change of light source illumination in a video collecting scene. The illumination distribution evaluation method can be combined with subsequent image enhancement processing such as histogram equalization, gamut mapping, gamma correction and the like, thereby achieving the effect of enhancing the video image quality.
Description
Technical field
The present invention relates to the video enhancement techniques field, specifically, relate to a kind of estimation method for video image illumination distribution.
Background technology
In video image enhancement process system, it is in order to strengthen the very important class technological means of picture quality that the Luminance Distribution of image is carried out processing such as equilibrium, mapping, correction.As common technologies such as histogram equalization, Color Gamut Mapping, gamma corrections.Yet image is carried out enhancement process, often need current pending video image brightness distribution is estimated.
Existing video image brightness distribution estimating technology mainly contains two big classes: a class is a local luminance distribution estimating technology, and another kind of is overall Luminance Distribution estimation technique.
Typical case's representative of local luminance distribution estimating is the self-adapting histogram equilibrium technology, be each local Luminance Distribution feature of adapting to image, the statistical picture brightness distributes partly, thereby calculate at the employed balance parameters of this image local, thereby reach adaptive histogram equalization target.
A kind of typical method that overall situation Luminance Distribution is estimated is as shown in Figure 1, image is divided into 5 zones once is numbered 0,1,2,3,4.Need to use earlier following luminance component formula commonly used for coloured image and calculate brightness:
Luminance＝0.27red+0.67green+0.06blue (1)
Wherein, red, green, blue represent coloured image redgreenblue component respectively.
After calculating the brightness value of each pixel according to formula (1), calculate the brightness arithmetic mean value of each subregion successively by subregion shown in Figure 1.A kind of result of calculation of possibility image brightness distribution as shown in Figure 2.According to brightness of image distribution of mean value result shown in Figure 2, it is bright roughly to judge the image upper left corner, and the lower right corner is darker.And then can be according to all fixed image enhancement technique, the concrete parameter when handling used subsequently of these data as histogram equalization, Color Gamut Mapping etc.
The technical scheme of carrying out the image brightness distribution correction according to the video image brightness distribution estimating of prior art has a basic defective.Specifically, the basis that prior art scheme Luminance Distribution is estimated is that its imaging process is shown in Fig. 3,4 according to calculating by the pixel brightness value that formula (1) carries out.In Fig. 3, after the light beam 102 that light source 101 sends incides the surface 105 of object scene 104, receive a part of incident light through object scene 104, diffuse reflection part incident light, last a branch of reflected light 103 enters video camera.In Fig. 4, object reflected light 103 projects on the video camera photoreceptor surface 202 through the light beam 201 behind the camera lens, and photoreceptor signal is through the digital image form 205 of the camera review treatment scheme final output in the 204 backs scene 104 of analog to digital conversion 203 and necessity.Object scene 104 final digital pictures are generally exported and are stored in the mode that three passages of each pixel R/G/B are represented by pixel.
By Fig. 3,4 as can be seen, the brightness value that formula (1) calculates has only reacted the luminance level of body surface folded light beam 103, but can not directly react the incident beam 102 that light source 101 projects object 104, the i.e. luminance state of illumination.In fact, with regard to general nature image imaging scene, the object scene surface reflectivity formula that spatially can acutely jump changes, and only can take place in the space slowly to change continuously and the light illumination in the reality is general.So the defective of existing conventional video image brightness distribution estimation technique is a spatially violent irregular variation of reaction material object scene surface reflectivity, and the variation spatially of object scene surface reflectivity is the object scene useful information that needs reservation record, and because light illumination changes the unbalanced part that is only the needs processing of the brightness of image that causes
Summary of the invention
The objective of the invention is to overcome the deficiency of existing video image brightness distribution estimating, a kind of Illumination Distribution and variation characteristic that can estimate the video image acquisition scene better is provided, thus the parameter adjustment that makes image enhancement processing subsequently estimation method for video image illumination distribution more accurately.
To achieve the above object of the invention, estimation method for video image illumination distribution of the present invention may further comprise the steps:
(1), calculates the corresponding brightness image of original image;
(2), use edge detection method to detect and draw out brightness edge of image image based on the image gradient value;
(3), on the basis of brightness image, calculate gradient image and gradient vector image;
(4), according to gradient image, calculate the gradient threshold values of gradient image; According to edge image and gradient threshold values the gradient vector image is carried out the gradient screening, if the locational gradient amplitude of the corresponding gradient image of certain pixel in the gradient vector image indicates this position to be present in the edge greater than gradient threshold values or edge image, then the gradient vector of certain pixel in the gradient vector image is set to zero, otherwise, keep;
(5), the gradient vector image after the screening carried out illumination image rebuild, obtain illumination image;
(6), carry out Illumination Distribution according to illumination image and estimate, obtain the video image Illumination Distribution and estimate.
In the present invention, through the gradient screening, remove the gradient vector influence of gradient amplitude greater than gradient threshold values or marginal existence part, through being carried out illumination image, rebuilds the gradient vector image after the screening, obtain illumination image, the influence that scene surface reflectivity acute variation is brought the image illumination distribution statistics be can effectively remove, thereby the feature and the changes in distribution of light illumination in the video image acquisition scene estimated exactly.Video image Illumination Distribution that the present invention obtains estimates to be the successive image enhancement process, as combinations such as histogram equalization, Color Gamut Mapping, gamma corrections, and then reaches the enhancing effect of video image quality.
Description of drawings
Fig. 1 is that prior art overall situation Luminance Distribution is estimated synoptic diagram;
Fig. 2 is the brightness distribution of mean value figure that subregion shown in Figure 1 calculates;
Fig. 3 is an object scene image-forming principle synoptic diagram;
Fig. 4 is a video image image-forming principle synoptic diagram;
Fig. 5 is a kind of embodiment process flow diagram of the present invention;
Fig. 6 is an integration reconstruct precedence diagram shown in Figure 5.
Embodiment
For understanding the present invention better, the present invention is more described in detail below in conjunction with embodiment.In the following description, when perhaps the detailed description of existing prior art can desalinate subject content of the present invention, these were described in here and will be left in the basket.
Fig. 1～4 are depicted as prior art, are described in background technology and illustrate, do not repeat them here.
Fig. 5 is a kind of embodiment process flow diagram of the present invention.In the present embodiment, may further comprise the steps:
Step ST1: each pixel of original image Img calculates brightness value according to following luminance component formula commonly used:
Luminance＝0.27red+0.67green+0.06blue (1)
Wherein, red, green, blue represent coloured image redgreenblue component respectively.
Obtain the corresponding brightness image I of original image mgL.
Step ST2: use edge detection method to detect and draw out edge image ImgE based on the image gradient value.It should be noted that present technique field personnel know detection and draw the method based on the image gradient value that edge image is not limited to mention here, can use numerous other tradition or emerging edge detection methods.
Step ST3: calculate its logarithm by pixel on the basis of brightness image I mgL, obtain logarithmic image ImgL ', computing formula is as follows:
ImgL′＝log?ImgL(2)
Wherein, ImgL ' logarithmic image value, ImgL is the brightness image value
Step ST4: after obtaining logarithmic image ImgL ', calculate its gradient vector image I mgGV and gradient image ImgG on its basis, specifically, and the usage space Difference Calculation, computing formula is as follows:
h(x，y)＝ImgL′(x+1，y)-ImgL′(x，y)；
V(x，y)＝ImgL′(x，y+1)-ImgL′(x，y) (3)
Wherein (x, the y) volume coordinate of presentation video, ImgGV (x, y)=(h (x, y), V (x, y)) constitutes the gradient vector of this location of pixels, and (x y) is the gradient amplitude value of this location of pixels to ImgG, from obtaining gradient vector image I mgGV and gradient image ImgG.
Step ST5: according to gradient image ImgG compute gradient threshold values GradT, concrete grammar is:
GradMax＝max(ImgG)；
GradMin＝min(imgG)； (4)
GradT＝GradMin+(GradMax-GradMin)*GradS
Wherein, GradS is experience data, and in this enforcement, value is 0.3.GradMax represents the greatest gradient range value among the gradient image ImgG, and GradMin represents the minimal gradient range value among the gradient image ImgG
On the basis of edge image ImgE and gradient threshold values GradT, carry out the gradient Screening Treatment.Specifically, be exactly by the pixel juggling, if (x y) indicates this position of this location of pixels to be present in the edge greater than gradient threshold values GradT or edge image ImgE to the gradient amplitude ImgG of gradient image ImgG on each location of pixels, and then this position gradient vector is set to zero.Concrete computing formula is as follows:
Step ST6: on the basis of the gradient vector image I mgGV ' after the screening, carry out the logarithm illumination image ImgIL of integration reconstituting initial image.Fig. 6 is an integration reconstruct precedence diagram shown in Figure 5, and among the figure, P11 presentation video top left corner pixel once is P12 line by line, and P13 etc. are P21 by leu, P31 etc.Being reconstructed is the value of the initial value use of P11 position from logarithmic image ImgL ' correspondence position P11, and the integration reconstruction calculations process of concrete illumination logarithmic image ImgIL is:
ImgIL(1，1)＝ImgL′(1，1)；
ImgIL(x+1，1)＝ImgIL(x，1)-ImgGV′(x，1)
v； (6)
ImgIL(x，y+1)＝ImgIL(x，y)-ImgGV′(x，y)
h；
Wherein, ImgIL (x, y) expression illumination logarithmic image in the position (x, the y) value on, ImgL ' (1,1) expression logarithmic image in the position (x, the y) value on, ImgGV ' (x, y)
vExpression gradient vector image in the position (x, vertical component y), ImgGV ' (x, y)
h(concrete integration reconstruction calculations line is that zigzag shown in Figure 6 is lined by line scan to expression gradient vector image for x, horizontal component y) in the position.
Step ST7: after obtaining illumination logarithmic image ImgIL, need carry out the inverse operation of formula (2) and calculate the final illumination image ImgI of original image, concrete computing method are:
ImgI＝log′ImgIL (7)
Wherein, ImgI illumination image value, ImgIL is an illumination logarithmic image value.
Step ST8: after obtaining illumination image ImgI, carry out Illumination Distribution and estimate, obtain the video image Illumination Distribution and estimate.The image illumination distribution estimating can adopt existing multiple statistical method to carry out.
In the present embodiment, by calculating edge image, the logarithm gradient image of original image, remove the graded that gradient amplitude value in the logarithm gradient image is higher than threshold values or has the location of pixels at edge effectively in conjunction with the calculating of gradient threshold values, process integration reconstruct and inverse logarithm calculate the illumination image of original image subsequently, carrying out Illumination Distribution then estimates, can be the successive image enhancement process, as combinations such as histogram equalization, Color Gamut Mapping, gamma corrections, and then reach the enhancing effect of video image quality.
Although above the illustrative embodiment of the present invention is described; but should be understood that; the invention is not restricted to the scope of embodiment; to those skilled in the art; as long as various variations appended claim limit and the spirit and scope of the present invention determined in; these variations are conspicuous, and all utilize innovation and creation that the present invention conceives all at the row of protection.
Claims (3)
1, a kind of estimation method for video image illumination distribution is characterized in that, may further comprise the steps:
(1), calculates the corresponding brightness image of original image;
(2), use edge detection method to detect and draw out brightness edge of image image based on the image gradient value;
(3), on the basis of brightness image, calculate gradient image and gradient vector image;
(4), according to gradient image, calculate the gradient threshold values of gradient image; According to edge image and gradient threshold values the gradient vector image is carried out the gradient screening, if the locational gradient amplitude of the corresponding gradient image of certain pixel in the gradient vector image indicates this position to be present in the edge greater than gradient threshold values or edge image, then the gradient vector of certain pixel in the gradient vector image is set to zero, otherwise, keep;
(5), the gradient vector image after the screening carried out illumination image rebuild, obtain illumination image;
(6), carry out Illumination Distribution according to illumination image and estimate, obtain the video image Illumination Distribution and estimate.
2, estimation method for video image illumination distribution according to claim 1 is characterized in that, step (3) is described to calculate gradient image and the gradient vector image step is:
(31), on the basis of brightness image, calculate its logarithm by pixel, obtain logarithmic image, computing formula is as follows:
ImgL′＝log?ImgL
Wherein, ImgL ' logarithmic image value, ImgL is the brightness image value
(32), obtain logarithmic image after, the usage space Difference Calculation goes out its gradient vector image and gradient image, computing formula is as follows:
h(x，y)＝ImgL′(x+1，y)-ImgL′(x，y)；
V(x，y)＝ImgL′(x，y+1)-ImgL′(x，y)
Wherein (x, the y) volume coordinate of presentation video, ImgGV (x, y)=(h (x, y), V (x, y)) constitutes the gradient vector of this location of pixels, and (x y) is the gradient amplitude value of this location of pixels, from obtaining gradient vector image and gradient image to ImgG.
3, estimation method for video image illumination distribution according to claim 2 is characterized in that, step (5) is described carries out the illumination image reconstruction to the gradient vector image after the screening, and step is:
(51), on the basis of the gradient vector image after the screening, carry out the logarithm illumination image of integration reconstituting initial image, the integration reconstruction calculations process of illumination logarithmic image is:
ImgIL(1，1)＝ImgL′(1，1)；
ImgIL(x+1，1)＝ImgIL(x，1)-ImgGV′(x，1)
v；
ImgIL(x，y+1)＝ImgIL(x，y)-ImgGV′(x，y)
h；
Wherein, ImgIL (x, y) expression illumination logarithmic image in the position (x, the y) value on, ImgL ' (1,1) expression logarithmic image in the position (x, the y) value on, ImgGV ' (x, y)
vExpression gradient vector image in the position (x, vertical component y), ImgGV ' (x, y)
hExpression gradient vector image is (x, horizontal component y) in the position.
(52), after obtaining the illumination logarithmic image, need carry out inverse operation and calculate the final illumination image of original image, concrete computing method are:
ImgI＝log′ImgIL
Wherein, ImgI illumination image value, ImgIL is an illumination logarithmic image value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008101478621A CN101510306B (en) | 2008-12-15 | 2008-12-15 | Estimation method for video image illumination distribution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008101478621A CN101510306B (en) | 2008-12-15 | 2008-12-15 | Estimation method for video image illumination distribution |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101510306A true CN101510306A (en) | 2009-08-19 |
| CN101510306B CN101510306B (en) | 2010-12-22 |
Family
ID=41002698
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008101478621A Expired - Fee Related CN101510306B (en) | 2008-12-15 | 2008-12-15 | Estimation method for video image illumination distribution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101510306B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102456222A (en) * | 2010-10-29 | 2012-05-16 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for organized equalization in image |
| CN103593847A (en) * | 2013-11-25 | 2014-02-19 | 中国航天科工集团第三研究院第八三五七研究所 | Raindrop detection analysis method based on machine vision |
| CN108805220A (en) * | 2018-07-25 | 2018-11-13 | 广东奥普特科技股份有限公司 | A kind of Fast template matching algorithm based on gradient integral |
| CN109118541A (en) * | 2018-08-24 | 2019-01-01 | 浙江大丰实业股份有限公司 | Stage orients ride guide rail position analysis system |
| CN114565535A (en) * | 2022-03-02 | 2022-05-31 | 北京锐影医疗技术有限公司 | Image enhancement method and device based on adaptive gradient gamma correction |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1139669A1 (en) * | 2000-03-28 | 2001-10-04 | STMicroelectronics S.r.l. | Coprocessor for motion estimation in digitised video sequence encoders |
| CN100548049C (en) * | 2007-01-09 | 2009-10-07 | 浙江大学 | A kind of method of the H.264 fast motion estimation based on multi-reference frame |
| CN100592797C (en) * | 2007-09-14 | 2010-02-24 | 四川虹微技术有限公司 | Fast motion estimating method |
-
2008
- 2008-12-15 CN CN2008101478621A patent/CN101510306B/en not_active Expired - Fee Related
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102456222A (en) * | 2010-10-29 | 2012-05-16 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for organized equalization in image |
| CN102456222B (en) * | 2010-10-29 | 2013-12-04 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for organized equalization in image |
| CN103593847A (en) * | 2013-11-25 | 2014-02-19 | 中国航天科工集团第三研究院第八三五七研究所 | Raindrop detection analysis method based on machine vision |
| CN108805220A (en) * | 2018-07-25 | 2018-11-13 | 广东奥普特科技股份有限公司 | A kind of Fast template matching algorithm based on gradient integral |
| CN109118541A (en) * | 2018-08-24 | 2019-01-01 | 浙江大丰实业股份有限公司 | Stage orients ride guide rail position analysis system |
| CN109118541B (en) * | 2018-08-24 | 2022-04-19 | 浙江大丰实业股份有限公司 | Stage directional vehicle platform guide rail position analysis system |
| CN114565535A (en) * | 2022-03-02 | 2022-05-31 | 北京锐影医疗技术有限公司 | Image enhancement method and device based on adaptive gradient gamma correction |
| CN114565535B (en) * | 2022-03-02 | 2022-09-13 | 北京锐影医疗技术有限公司 | Image enhancement method and device based on adaptive gradient gamma correction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101510306B (en) | 2010-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ancuti et al. | I-HAZE: A dehazing benchmark with real hazy and haze-free indoor images | |
| Park et al. | Single image dehazing with image entropy and information fidelity | |
| US9852499B2 (en) | Automatic selection of optimum algorithms for high dynamic range image processing based on scene classification | |
| Tang et al. | Investigating haze-relevant features in a learning framework for image dehazing | |
| US9842382B2 (en) | Method and device for removing haze in single image | |
| CN106030653B (en) | For generating the image processing system and image processing method of high dynamic range images | |
| CN102006484B (en) | Image processing device, image processing method, and program | |
| CN107292830B (en) | Low-illumination image enhancement and evaluation method | |
| Peng et al. | Image haze removal using airlight white correction, local light filter, and aerial perspective prior | |
| CN109754377A (en) | A kind of more exposure image fusion methods | |
| CN108765342A (en) | A kind of underwater image restoration method based on improvement dark | |
| CN107451969A (en) | Image processing method, device, mobile terminal and computer-readable recording medium | |
| CN103606132A (en) | Multiframe digital image denoising method based on space domain and time domain combination filtering | |
| TWI518635B (en) | Image restoration method and image processing apparatus using the same | |
| CN103218778A (en) | Image and video processing method and device | |
| CN101626454B (en) | Method for intensifying video visibility | |
| CN110689490A (en) | Underwater image restoration method based on texture color features and optimized transmittance | |
| CN112785534A (en) | Ghost-removing multi-exposure image fusion method in dynamic scene | |
| CN101510306B (en) | Estimation method for video image illumination distribution | |
| CN105959510B (en) | A kind of video rapid defogging method | |
| CN112640413B (en) | Method for displaying a model of the surroundings, control device and vehicle | |
| CN104823437A (en) | Picture processing method and device | |
| TW201308251A (en) | Underwater image enhancement system | |
| US20190068891A1 (en) | Method and apparatus for rapid improvement of smog/low-light-level image using mapping table | |
| TWI376648B (en) | Method and device for keeping image background by multiple gauss models |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20101222 Termination date: 20151215 |
|
| EXPY | Termination of patent right or utility model |