Summary of the invention
Technical problem to be solved by this invention is to provide a kind of processing method of multi-view image, and it can accurately realize consistency of colour between viewpoint, and need not with reference to original visual point image when extracting watermark.
The present invention solves the problems of the technologies described above the technical scheme that adopts: a kind of processing method of multi-view image is characterized in that its processing procedure is: at first, adopt histogram matching to set up the color map relation of target image and source images; Secondly, the function representation form according to Gamma correction is embedded into watermarking images to be embedded in source images, obtains being embedded with the source images of watermark; Then, adopt the multiple view video coding device to carry out coding transmission to target image and the source images that is embedded with watermark, the source images that is embedded with watermark after target image after user side is to coding and coding is decoded, and obtains decoded target image and the decoded source images that is embedded with watermark; At last; each color component to each pixel in the decoded source images that is embedded with watermark carries out color correction; obtain the decoded correcting image of source images after color correction that is embedded with watermark; or extract the watermark image from the decoded source images that is embedded with watermark, whether protected by copyrightly then determine to be embedded with the source images of watermark according to the watermark image.
The processing method of above-mentioned a kind of multi-view image, it specifically comprises the following steps:
1. synchronization is defined as target image by a visual point image in the multi-view image of the parallel camera system shooting of many viewpoints, be designated as
(T)I
i(x
1, y
1), and other visual point image is defined as source images, be designated as
(S)I
i(x
1, y
1), wherein, i=1,2,3 represent respectively luminance component Y, the first chromatic component U and the second chromatic component V in three color components of YUV color space, (x
1, y
1) coordinate position of pixel in expression target image and source images, 1≤x
1≤ W, 1≤y
1≤ H, W represent the width of target image and source images, and H represents the height of target image and source images,
(T)I
i(x
1, y
1) expression
(T)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel,
(S)I
i(x
1, y
1) expression
(S)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel;
2. adopt histogram matching, set up target image
(T)T
i(x
1, y
1) in each color value and source images
(S)I
i(x
1, y
1) in the color map relation of corresponding color value, be designated as q=f
i(p), wherein, f
iThe function representation form of color map relation of i color component of () expression, p represent source images
(S)I
i(x
1, y
1) in color value, q represent source images
(S)I
i(x
1, y
1) in the color value of color value after color correction, 0≤p≤255,0≤q≤255;
3. according to the function representation form r=α * p of Gamma correction
β, with watermarking images to be embedded { P (x
2, y
2) be embedded into source images
(S)I
i(x
1, y
1) in, obtain being embedded with the source images of watermark, be designated as
Wherein, α and β be for controlling parameter, p represent source images
(S)I
i(x
1, y
1) in color value, r represent source images
(S)I
i(x
1, y
1) in the color value of color value after Gamma correction, 0≤p≤255,0≤r≤255,
Expression is embedded with the source images of watermark
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel, (x
2, y
2) expression watermarking images { P (x to be embedded
2, y
2) in the coordinate position of pixel, 1≤x
2≤ N, 1≤y
2≤ N, N represent watermarking images to be embedded { P (x
2, y
2) the horizontal and vertical size, P (x
2, y
2) expression watermarking images { P (x to be embedded
2, y
2) in coordinate position be (x
2, y
2) the watermark information of pixel;
4. adopt the multiple view video coding device to target image
(T)I
i(x
1, y
1) and be embedded with the source images of watermark
Carry out coding transmission, the source images that is embedded with watermark after the target image after user side is to coding and coding is decoded, and obtains decoded target image and the decoded source images that is embedded with watermark, and correspondence is designated as respectively
With
Wherein,
Represent decoded target image
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel,
Represent the decoded source images that is embedded with watermark
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel;
5. in different ways to the decoded source images that is embedded with watermark
Process, if select the color correction mode, execution in step 6., if select the copyright protection pattern, execution in step is 7.;
6. according to q=f
i(p), to the decoded source images that is embedded with watermark
In each color component of each pixel carry out color correction, obtain
Correcting image after color correction, be designated as
(C)I
i(x
1, y
1), will
(C)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel be designated as
(C)I
i(x
1, y
1),
7. according to r=α * p
β, from the decoded source images that is embedded with watermark
In extract the watermark image, be designated as
(Q)I
i(x
1, y
1), will
(Q)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel be designated as
(Q)I
i(x
1, y
1),
Then judge whether can from the watermark image
(Q)I
i(x
1, y
1) in observe the watermarking images { P (x of embedding
2, y
2) in the watermark information of each pixel, if can, determine to be embedded with the source images of watermark
Protected by copyright, otherwise, determine to be embedded with the source images of watermark
Not protected by copyright.
Described step detailed process 2. is:
2.-1, calculate target image
(T)I
i(x
1, y
1) the histogram of i color component, be designated as
(T)H
i(k) | 0≤k≤255}, then calculate
(T)H
i(k) | the accumulation histogram of 0≤k≤255}, be designated as
(T)C
i(k) | 0≤k≤255},
Wherein,
(T)H
i(k) expression target image
(T)I
i(x
1, y
1) the color value of i color component be the number of all pixels of k,
(T)H
i(j) expression target image
(T)I
i(x
1, y
1) the color value of i color component be the number of all pixels of j,
(T)C
i(k) expression target image
(T)I
i(x
1, y
1) the color value of i color component be not more than the accumulative total number of all pixels of k;
2.-2, calculate source images
(S)I
i(x
1, y
1) the histogram of i color component, be designated as
(S)H
i(k) | 0≤k≤255}, then calculate
(S)H
i(k) | the accumulation histogram of 0≤k≤255}, be designated as
(S)C
i(k) | 0≤k≤255},
Wherein,
(S)H
i(k) expression source images
(S)I
i(x
1, y
1) the color value of i color component be the number of all pixels of k,
(S)H
i(j) expression source images
(S)I
i(x
1, y
1) the color value of i color component be the number of all pixels of j,
(S)C
i(k) expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than the accumulative total number of all pixels of k;
2.-3, get source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
0The accumulative total number of all pixels
(S)C
i(p
0), then calculate
(S)C
i(p
0) and target image
(T)I
i(x
1, y
1) the color value of i color component be not more than the accumulative total number of all pixels of q
(T)C
i(q) distance between is designated as
(S)C
i(p
0)-
(T)C
i(q) |, then basis
(S)C
i(p
0)-
(T)C
i(q) | obtain source images
(S)I
i(x
1, y
1) in color value p
0With target image
(T)I
i(x
1, y
1) in color value q
0The color map relation, be designated as q
0=f
i(p
0),
Wherein, 0≤p
0≤ 255,0≤q
0≤ 255,0≤q≤255, " || " is the symbol that takes absolute value,
Expression makes
(S)C
i(p
0)-
(T)C
i(q) | the q value of value minimum;
2.-4, choose 255 and p
0Different color value is designated as respectively p
1, p
2..., p
j′..., p
255, then according to the operating process of step 2.-3, obtain respectively in an identical manner source images
(S)I
i(x
1, y
1) in color value p
1, p
2..., p
j′..., p
255With target image
(T)I
i(x
1, y
1) in color value q
1, q
2..., q
j,, q
255Color map relation one to one, correspondence is designated as q respectively
1=f
i(p
1), q
2=f
i(p
2) ..., q
j′=f
i(p
j′) ..., q
255=f
i(p
255),
Wherein, 1≤j'≤255,0≤q≤255,0≤p
1, p
2..., p
j'..., p
255≤ 255,0≤q
1, q
2..., q
j′..., q
255≤ 255, q
1, q
2..., q
j′..., q
255Value and q
0Value different,
(S)C
i(p
1) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
1The accumulative total number of all pixels,
(S)C
i(p
2) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
2The accumulative total number of all pixels,
(S)C
i(p
j′) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
j'The accumulative total number of all pixels,
(S)C
i(p
255) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
255The accumulative total number of all pixels.
Described step detailed process 3. is:
3.-1, according to r=α * p
β, obtain watermarked optimum density value, be designated as p
*,
Wherein,
For getting gradient function,
Expression makes
Maximum p value;
3.-2, according to watermarked optimum density value p
*, obtain source images
(S)I
i(x
1, y
1) in watermarked original position, be designated as
Wherein,
1≤x
0≤ W-N+1,1≤y
0≤ H-N+1,
(S)I
i(x
1, y
1) the expression source images
(S)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of the 1st color component of pixel, " || " is the symbol that takes absolute value,
Expression makes
Maximum (x
0, y
0);
3.-3, with source images
(S)I
i(x
1, y
1) in abscissa exist
In interval and ordinate exist
The current pixel of processing in interval is defined as current pixel point;
3. the coordinate position of-4, supposing current pixel point is (x
1', y
1'), then judge watermarking images to be embedded { P (x
2, y
2) in coordinate position be
The watermark information of pixel whether be 1, if so, revise the color value of the 1st color component of current pixel point, the color value of the 1st color component of amended current pixel point is designated as
Otherwise, directly with the color value of the 1st color component of the current pixel point color value as the 1st color component of amended current pixel point, the color value of the 1st color component of amended current pixel point is designated as
Wherein,
(S)I
i(x
1', y
1') the expression source images
(S)I
i(x
1, y
1) in coordinate position be (x
1', y
1') the color value of the 1st color component of pixel, Δ is intensity factor;
3.-5, with source images
(S)I
i(x
1, y
1) in abscissa exist
In interval and ordinate exist
Then next pending pixel in interval returns to step 3.-4 and continues to carry out as current pixel point, until source images
(S)I
i(x
1, y
1) in abscissa exist
In interval and ordinate exist
All pixels in interval are disposed, and obtain being embedded with the source images of watermark, are designated as
Wherein,
Expression is embedded with the source images of watermark
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel.
Described step is got α=1, β=0.2 in 3..
Get Δ=3 in described step 3.-3.
Compared with prior art, the invention has the advantages that:
1) the inventive method reaches the purpose of color correction and copyright protection by the color map relation that changes source images and target image; adopt histogram matching to realize accurate color correction; make solid colour between viewpoint; and adopt Gamma correction to strengthen the watermark strength of embedding; its watermark embedding and extracting method are very simple; and need not when extracting watermark with reference to original visual point image, can be effectively applied in various multi-view video systems.
2) the inventive method has designed color correction and the copyright protection dual mode is processed the decoded source images that is embedded with watermark; like this when guaranteeing multi-view image color correction accuracy; can effectively protect the copyright information of multi-view image, improve the validity of multi-view video system.
Description of drawings
Fig. 1 is the FB(flow block) of the inventive method;
Fig. 2 a is the target image of " flamenco1 " many viewpoints test set;
Fig. 2 b is the source images of " flamenco1 " many viewpoints test set;
Fig. 2 c is the target image of " objects2 " many viewpoints test set;
Fig. 2 d is the source images of " objects2 " many viewpoints test set;
Fig. 3 a " University Of Ningbo " watermarking images for embedding;
Fig. 3 b " Information Institute " watermarking images for embedding;
Fig. 3 c " image processing " watermarking images for embedding;
Fig. 3 d " electronic information " watermarking images for embedding;
Fig. 3 e " one two three four " watermarking images for embedding;
Fig. 4 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 a obtains through the inventive method;
Fig. 4 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 a obtains through the inventive method;
Fig. 4 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 a obtains through the inventive method;
Fig. 4 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 a obtains through the inventive method;
Fig. 5 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 b obtains through the inventive method;
Fig. 5 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 b obtains through the inventive method;
Fig. 5 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 b obtains through the inventive method;
Fig. 5 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 b obtains through the inventive method;
Fig. 6 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 c obtains through the inventive method;
Fig. 6 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 c obtains through the inventive method;
Fig. 6 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 c obtains through the inventive method;
Fig. 6 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 c obtains through the inventive method;
Fig. 7 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 d obtains through the inventive method;
Fig. 7 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 d obtains through the inventive method;
Fig. 7 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 d obtains through the inventive method;
Fig. 7 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 d obtains through the inventive method;
Fig. 8 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 e obtains through the inventive method;
Fig. 8 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 e obtains through the inventive method;
Fig. 8 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 e obtains through the inventive method;
Fig. 8 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 e obtains through the inventive method.
Embodiment
Embodiment is described in further detail the present invention below in conjunction with accompanying drawing.
The processing method of a kind of multi-view image that the present invention proposes, its flow process as shown in Figure 1, its processing procedure is: at first, adopt histogram matching to set up the color map relation of target image and source images; Secondly, the function representation form according to Gamma correction is embedded into watermarking images to be embedded in source images, obtains being embedded with the source images of watermark; Then, adopt the multiple view video coding device to carry out coding transmission to target image and the source images that is embedded with watermark, the source images that is embedded with watermark after target image after user side is to coding and coding is decoded, and obtains decoded target image and the decoded source images that is embedded with watermark; At last; each color component to each pixel in the decoded source images that is embedded with watermark carries out color correction; obtain the decoded correcting image of source images after color correction that is embedded with watermark; or extract the watermark image from the decoded source images that is embedded with watermark, whether protected by copyrightly then determine to be embedded with the source images of watermark according to the watermark image.
The processing method of multi-view image of the present invention, it specifically comprises the following steps:
1. synchronization is defined as target image by a visual point image in the multi-view image of the parallel camera system shooting of many viewpoints, be designated as
(T)I
i(x
1, y
1), and other visual point image is defined as source images, be designated as
(S)I
i(x
1, y
1), wherein, i=1,2,3 represent respectively luminance component Y, the first chromatic component U and the second chromatic component V in three color components of YUV color space, (x
1, y
1) coordinate position of pixel in expression target image and source images, 1≤x
1≤ W, 1≤y
1≤ H, W represent the width of target image and source images, and H represents the height of target image and source images,
(T)I
i(x
1, y
1) expression
(T)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel,
(S)I
i(x
1, y
1) expression
(S)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel.
In the present embodiment, adopt multi-view point video test set " flamencol " that Japanese KDDI company provides and " objects2 " as original multi-viewpoint video image.Fig. 2 a and Fig. 2 b are respectively target image and the source images of " flamenco1 " many viewpoints test set, and Fig. 2 c and Fig. 2 d are respectively target image and the source images of " objects2 " many viewpoints test set, and the size of target image and source images is 320 * 240.Can find out from Fig. 2 a, Fig. 2 b, Fig. 2 c and Fig. 2 d, the color appearance of target image and source images is obviously inconsistent, and therefore it is carried out color correction just seems very necessary.
2. can reflect the Distribution Statistics information of color of image due to histogram, so the present invention adopts histogram matching, set up target image
(T)I
i(x
1, y
1) in each color value and source images
(S)I
i(x
1, y
1) in the color map relation of corresponding color value, be designated as q=f
i(p), wherein, f
iThe function representation form of color map relation of i color component of () expression, p represent source images
(S)I
i(x
1, y
1) in color value, q represent source images
(S)I
i(x
1, y
1) in the color value of color value after color correction, i.e. target image corresponding to expression
(T)I
i(x
1, y
1) in color value, 0≤p≤255,0≤q≤255.
In this specific embodiment, step detailed process 2. is:
2.-1, calculate target image
(T)I
i(x
1, y
1) the histogram of i color component, be designated as
(T)H
i(k) | 0≤k≤255}, then calculate
(T)H
i(k) | the accumulation histogram of 0≤k≤255}, be designated as
(T)C
i(k) | 0≤k≤255},
Wherein,
(T)H
i(k) expression target image
(T)I
i(x
1, y
1) the color value of i color component be the number of all pixels of k,
(T)H
i(j) expression target image
(T)I
i(x
1, y
1) the color value of i color component be the number of all pixels of j,
(T)C
i(k) expression target image
(T)I
i(x
1, y
1) the color value of i color component be not more than the accumulative total number of all pixels of k.
2.-2, calculate source images
(S)I
i(x
1, y
1) the histogram of i color component, be designated as
(S)H
i(k) | 0≤k≤255}, then calculate
(S)H
i(k) | the accumulation histogram of 0≤k≤255}, be designated as
(S)C
i(k) | 0≤k≤255},
Wherein,
(S)H
i(k) expression source images
(S)I
i(x
1, y
1) the color value of i color component be the number of all pixels of k,
(S)H
i(j) expression source images
(S)I
i(x
1, y
1) the color value of i color component be the number of all pixels of j,
(S)C
i(k) expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than the accumulative total number of all pixels of k.
2.-3, get source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
0The accumulative total number of all pixels
(S)C
i(p
0), then calculate
(S)C
i(p
0) and target image
(T)I
i(x
1, y
1) the color value of i color component be not more than the accumulative total number of all pixels of q
(T)C
i(q) distance between is designated as
(S)C
i(p
0)-
(T)C
i(q) |, then basis
(S)C
i(p
0)-
(T)C
i(q) | obtain source images
(S)I
i(x
1, y
1) in color value p
0With target image
(T)I
i(x
1, y
1) in color value q
0The color map relation, be designated as q
0=f
i(p
0),
Wherein, 0≤p
0≤ 255,0≤q
0≤ 255,0≤q≤255, " || " is the symbol that takes absolute value,
Expression makes
(S)C
i(p
0)-
(T)C
i(q) | the q value of value minimum.
2.-4, choose 255 and p
0Different color value is designated as respectively p
1, p
2..., p
j′..., p
255, then according to the operating process of step 2.-3, obtain respectively in an identical manner source images
(S)I
i(x
1, y
1) in color value p
1, p
2..., p
j′..., p
255With target image
(T)I
i(x
1, y
1) in color value q
1, q
2..., q
j′..., q
255Color map relation one to one, correspondence is designated as q respectively
1=f
i(p
1), q
2=f
i(p
2) ..., q
j′=f
i(p
j′) ..., q
255=f
i(p
255),
Wherein, 1≤j'≤255,0≤q≤255,0≤p
1, p
2..., p
j'..., p
255≤ 255,0≤q
1, q
2..., q
j′..., q
255≤ 255, q
1, q
2..., q
j′..., q
255Value and q
0Value different,
(S)C
i(p
1) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
1The accumulative total number of all pixels,
(S)C
i(p
2) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
2The accumulative total number of all pixels,
(S)C
i(p
j′) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
j'The accumulative total number of all pixels,
(S)C
i(p
255) the expression source images
(S)I
i(x
1, y
1) the color value of i color component be not more than p
255The accumulative total number of all pixels.
3. because Gamma correction can expand to less misalignment in larger color gamut, so the present invention is according to the function representation form r=α * p of Gamma correction
β, with watermarking images to be embedded { P (x
2, y
2) be embedded into source images
(S)I
i(x
1, y
1) in, obtain being embedded with the source images of watermark, be designated as
Wherein, α and β in the present embodiment, get α=1, β=0.2 for controlling parameter, p represent source images
(S)I
i(x
1, y
1) in color value, r represent source images
(S)I
i(x
1, y
1) in the color value of color value after Gamma correction, 0≤p≤255,0≤r≤255,
Expression is embedded with the source images of watermark
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel, (x
2, y
2) expression watermarking images { P (x to be embedded
2, y
2) in the coordinate position of pixel, 1≤x
2≤ N, 1≤y
2≤ N, N represent watermarking images to be embedded { P (x
2, y
2) the horizontal and vertical size, P (x
2, y
2) expression watermarking images { P (x to be embedded
2, y
2) in coordinate position be (x
2, y
2) the watermark information of pixel.
In this specific embodiment, step detailed process 3. is:
3.-1, according to r=α * p
β, obtain watermarked optimum density value, be designated as p
*,
Wherein,
For getting gradient function,
Expression makes
Maximum p value.
3.-2, according to watermarked optimum density value p
*, obtain source images
(S)I
i(x
1, y
1) in watermarked original position, be designated as
Wherein,
1≤x
0≤ W-N+1,1≤y
0≤ H-N+1,
(S)I
i(x
1, y
1) the expression source images
(S)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of the 1st color component of pixel, " || " is the symbol that takes absolute value,
Expression makes
Maximum (x
0, y
0).
3.-3, with source images
(S)I
i(x
1, y
1) in from original position
Beginning, abscissa exists
In interval and ordinate exist
The current pixel of processing in interval is defined as current pixel point.
3. the coordinate position of-4, supposing current pixel point is (x
1', y
1'), then judge watermarking images to be embedded { P (x
2, y
2) in coordinate position be
The watermark information of pixel whether be 1, if so, revise the color value of the 1st color component of current pixel point, the color value of the 1st color component of amended current pixel point is designated as
Otherwise, directly with the color value of the 1st color component of the current pixel point color value as the 1st color component of amended current pixel point, the color value of the 1st color component of amended current pixel point is designated as
Wherein,
(S)I
i(x
1', y
1') the expression source images
(S)I
i(x
1, y
1) in coordinate position be (x
1', y
1') the color value of the 1st color component of pixel, Δ is intensity factor, in the present embodiment, gets Δ=3.
3.-5, with source images
(S)I
i(x
1, y
1) in abscissa exist
In interval and ordinate exist
Then next pending pixel in interval returns to step 3.-4 and continues to carry out as current pixel point, until source images
(S)I
i(x
1, y
1) in abscissa exist
In interval and ordinate exist
All pixels in interval are disposed, and obtain being embedded with the source images of watermark, are designated as
Wherein,
Expression is embedded with the source images of watermark
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel.
4. adopt the multiple view video coding device to target image
(T)I
i(x
1, y
1) and be embedded with the source images of watermark
Carry out coding transmission, the source images that is embedded with watermark after the target image after user side is to coding and coding is decoded, and obtains decoded target image and the decoded source images that is embedded with watermark, and correspondence is designated as respectively
With
Wherein,
Represent decoded target image
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel,
Represent the decoded source images that is embedded with watermark
Middle coordinate position is (x
1, y
1) the color value of i color component of pixel.
In the present embodiment, the multiple view video coding device adopts known HBP coded prediction structure.
5. in different ways to the decoded source images that is embedded with watermark
Process, if select the color correction mode, execution in step 6., if select the copyright protection pattern, execution in step is 7..
6. according to q=f
i(p), to the decoded source images that is embedded with watermark
In each color component of each pixel carry out color correction, obtain
Correcting image after color correction, be designated as
(C)I
i(x
1, y
1), will
(C)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel be designated as
(C)I
i(x
1, y
1),
7. according to r=α * p
β, from the decoded source images that is embedded with watermark
In extract the watermark image, be designated as
(Q)I
i(x
1, y
1), will
(Q)I
i(x
1, y
1) in coordinate position be (x
1, y
1) the color value of i color component of pixel be designated as
(Q)I
i(x
1, y
1),
Then judge whether can from the watermark image
(Q)I
i(x
1, y
1) in observe the watermarking images { P (x of embedding
2, y
2) in the watermark information of each pixel, if can, determine to be embedded with the source images of watermark
Protected by copyright, otherwise, determine to be embedded with the source images of watermark
Not protected by copyright.
Below with regard to the inventive method, the subjective performance that " flamenco1 " and " objects2 " two groups of multi-view point video test sets carry out color correction and watermark extracting is compared.
Fig. 3 a, Fig. 3 b, Fig. 3 c, Fig. 3 d and Fig. 3 e have provided respectively five groups of binary bitmap to be embedded, and each resolution of organizing watermarking images is 80 * 80.Fig. 4 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 a obtains through the inventive method, Fig. 4 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 a obtains through the inventive method, Fig. 5 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 b obtains through the inventive method, Fig. 5 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 b obtains through the inventive method, Fig. 6 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 c obtains through the inventive method, Fig. 6 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 c obtains through the inventive method, Fig. 7 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 d obtains through the inventive method, Fig. 7 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 d obtains through the inventive method, Fig. 8 a is the color correction image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 e obtains through the inventive method, Fig. 8 c is the color correction image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 e obtains through the inventive method, from Fig. 4 a, Fig. 4 c, Fig. 5 a, Fig. 5 c, Fig. 6 a, Fig. 6 c, Fig. 7 a, Fig. 7 c, Fig. 8 a, the subjective effect of the image shown in Fig. 8 c can be found out, compare with the target image of " objects2 " shown in Fig. 2 c with the target image of " flamenco1 " shown in Fig. 2 a, color appearance and the target image of the color correction image that employing the inventive method obtains are very approaching, and can not perceive the watermark information of embedding from the color correction image, the watermark embed process that the inventive method is described is effective.
Fig. 4 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 a obtains through the inventive method, Fig. 4 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 a obtains through the inventive method, Fig. 5 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 b obtains through the inventive method, Fig. 5 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 b obtains through the inventive method, Fig. 6 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 c obtains through the inventive method, Fig. 6 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 c obtains through the inventive method, Fig. 7 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 d obtains through the inventive method, Fig. 7 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 d obtains through the inventive method, Fig. 8 b is the watermark image that the image after the watermarking images shown in Fig. 2 b embedding Fig. 3 e obtains through the inventive method, Fig. 8 d is the watermark image that the image after the watermarking images shown in Fig. 2 d embedding Fig. 3 e obtains through the inventive method, from Fig. 4 b, Fig. 4 d, Fig. 5 b, Fig. 5 d, Fig. 6 b, Fig. 6 d, Fig. 7 b, Fig. 7 d, Fig. 8 b, watermark image shown in Fig. 8 d can be observed the watermarking images of embedding at an easy rate, the watermark extraction process that the inventive method is described is effective.