WO2008089683A1 - Procédé, système et dispositif de correction de gamma - Google Patents

Procédé, système et dispositif de correction de gamma Download PDF

Info

Publication number
WO2008089683A1
WO2008089683A1 PCT/CN2008/070122 CN2008070122W WO2008089683A1 WO 2008089683 A1 WO2008089683 A1 WO 2008089683A1 CN 2008070122 W CN2008070122 W CN 2008070122W WO 2008089683 A1 WO2008089683 A1 WO 2008089683A1
Authority
WO
WIPO (PCT)
Prior art keywords
vaps
determining
vcsptn
color space
mbc
Prior art date
Application number
PCT/CN2008/070122
Other languages
English (en)
Chinese (zh)
Inventor
Zhong Luo
Original Assignee
Huawei Technologies Co., Ltd.
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co., Ltd. filed Critical Huawei Technologies Co., Ltd.
Publication of WO2008089683A1 publication Critical patent/WO2008089683A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N9/00—Details of colour television systems
    • H04N9/64—Circuits for processing colour signals
    • H04N9/68—Circuits for processing colour signals for controlling the amplitude of colour signals, e.g. automatic chroma control circuits
    • H04N9/69—Circuits for processing colour signals for controlling the amplitude of colour signals, e.g. automatic chroma control circuits for modifying the colour signals by gamma correction

Definitions

  • Gamma correction method Gamma correction method, gamma correction system, and gamma correction device
  • the present invention relates to the field of communications, and more particularly to gamma correction techniques for video communications. Background technique
  • video communication-based services also referred to as video communication services
  • video communication services have also been widely used.
  • video conferencing and video telephony services are becoming next generation networks
  • the video communication system architecture for implementing the video communication service is as shown in FIG. 1, and includes a video input device, a first color conversion module, a first signal processing module, a network, a second signal processing module, a second color conversion module, and a video output device.
  • the working principle is as follows: the video signal is captured by the video input device as an RGB optical signal to form an RGB electrical signal, and then the color conversion is performed by the first color conversion module, and the RGB electrical signal is converted into a YUV electrical signal, and then passed.
  • the first signal processing module performs signal processing (compression coding, etc.) on the YUV electrical signal, and then transmits the signal to the second signal processing module through the network for signal processing (decompression decoding, etc.), and then passes through the second color conversion module.
  • the inverse color transformation is performed to convert the YUV signal into an RGB electrical signal, which is then converted to an optical signal when displayed on the video output device.
  • the video input device and the video display device at both ends are located in the RGB color space, and various signal processing (compression coding, decompression decoding, etc.) and network transmission process in the middle are performed in the YUV color space.
  • the RGB color space is generated by a linear combination of three primary colors of red, green, and blue (RGB) (i.e., different ratios are added and mixed).
  • RGB red, green, and blue
  • each color can represent a point in three-dimensional space if its coordinates are represented by RGB components.
  • the YUV color space is introduced for improving the efficiency of signal processing in various signal processing and network transmission processes in the middle, which is a color space commonly used in video processing (there are many Variants such as YCC/YCbCr space, etc., whose expression is as shown in the formula [1]: y 0.299 0.587 0.114 r
  • the specific brightness or chrominance signal levels can be different in each processing step of the video signal, such as 256, 64, etc., but can be changed by dividing the brightness or chrominance signal of each processing step to the highest level.
  • the color space is usually represented by a normalized representation.
  • r, g, and b all take positive values, after using the normalized representation, 0 r, g, bl is required, that is, the absolute values of the r, g, and b components are not greater than 1; In space, the requirement - 1 ⁇ y, u, v ⁇ 1.
  • RGB color space if normalized, the RGB color space is a unit cube [0, 1] X [0, 1] X [0, 1], or simply written as [0, 1] 3 , where The multiplication sign "X" represents the Cartesian Product of the set.
  • the resulting normalized cube is called the RGB Unit Cube (RGBUC). As shown in Figure 3.
  • the impact is the most In addition to the QoS (Quality of Service) parameters of the network (including packet loss, delay, jitter, R-factor, etc.), there are also distortions of the luminance signal caused by the Gamma characteristics of each link. (Distortion) factor.
  • the QoS parameters of the network including packet loss, delay, jitter, R-factor, etc.
  • the pre-processing post-processing
  • gamma correction processing is required to Make the final input-output relationship of the signal a linear relationship. In this way, the video/still image captured by the video input device (such as the camera/camera) is displayed on the display device to achieve a high quality display effect and a good user experience.
  • the Gamma characteristic refers to a non-linear relationship between the luminance signal input-output relationship of a certain link. After the distortion of the Gamma nonlinear link, the luminance signal is increased according to the power function.
  • g r , g g , g b represent the Gamma characteristics of the 1, B, and G components, respectively.
  • r r represents the original R component signal
  • the subscript r raw, representing the original
  • r d represents the R component signal after the gamma distortion
  • the subscript d distorted, indicating distortion.
  • Figure 4 shows the Gamma characteristic curve of a video input device, where the red, green, and blue curves represent the Gamma characteristic curves corresponding to the functions gr, gg, and gb, respectively.
  • the prior art related to the present invention gives a working principle of gamma correction for a captured signal based on the RGB color space through a Gamma correction module built in a part of the high-end video input device, as shown in FIG.
  • the link adds a gamma correction module to perform gamma correction on the input luminance signal, and the gamma characteristic is expressed as Gc(.), through the Gamma school.
  • the input-output relationship is a linear relationship.
  • Embodiments of the present invention provide a gamma correction method, a gamma correction system, and a gamma correction device, by which gamma correction can be performed in various color spaces used in an intermediate processing section in a video communication process.
  • An embodiment of the present invention provides a gamma correction method for video communication, including: dividing a VAPS into a plurality according to a minimum inclusion cuboid MBC of a virtual color space allowable point set VAPS in a virtual color space to form a plurality of VCSPTN division;
  • gamma correction is performed on each of the obtained virtual color space vector signal samples.
  • An embodiment of the present invention further provides a gamma correction system for video communication, comprising: a cell determination unit, configured to select a minimum inclusion box MBC of a point set VAPS according to a virtual color space in a virtual color space, to a VAPS Performing a division to form a plurality of VCSPTN cells;
  • a linear gamma correction function determining unit configured to construct a linear gamma correction function on each VCSPTN partition; a cell matching unit, configured to obtain a virtual color space vector signal sample value corresponding to each original RGB vector electrical signal value included in the video information, and determine a VCSPTN cell to which it belongs; each obtained The virtual color space vector signal samples are gamma corrected.
  • An embodiment of the present invention further provides a gamma correction device, including:
  • a cell determining unit configured to divide the VAPS into different VCSPTN cells according to a minimum inclusion box MBC of the virtual color space allowable point set VAPS in the virtual color space;
  • a linear gamma correction function determining unit for constructing a linear gamma correction function on each VCSPTN partition.
  • the present invention first divides the VAPS according to the minimum inclusion box MBC of the virtual color space allowable point set VAPS in the virtual color space to form a plurality of VCSPTN points. And then construct a linear gamma correction function on each VCSPTN partition; and then obtain a virtual color space vector signal sample value corresponding to each original RGB vector electrical signal included in the video information, and determine the location thereof Each of the virtual color space vector signal values that are attributed is subjected to gamma correction.
  • gamma correction can be implemented in the color space used in the intermediate processing link in the video communication process, thereby expanding the application range of the gamma correction, and solving the current gamma correction cannot be performed in the intermediate processing of the video communication. Puzzle. DRAWINGS
  • FIG. 1 is a structural diagram of a video communication system provided by the background art
  • 2 is a schematic diagram showing the principle of transformation from RGB color space to YUV color space provided by the background art
  • 3 is an RGB unit cube obtained by normalizing the RGB color space provided by the background art
  • FIG. 5 is a schematic diagram of a correction principle of Gamma distortion introduced by a video input device provided by the background art
  • FIG. 6 is a schematic diagram of a VCS color space introduced by the present invention.
  • FIG. 7 is a schematic diagram showing the relationship between a VAPS and a minimum including a cuboid in a VCS according to the present invention
  • FIG. 8 is a flowchart of a first embodiment provided by the present invention.
  • FIG. 9 is a schematic diagram showing the principle of a VAPS set in a first embodiment provided by the present invention
  • FIG. 10 is a schematic diagram showing the six faces of a VCSPTN cell in the first embodiment of the present invention.
  • FIG. 11 is a diagram showing a correspondence relationship between a VCSPTN cell and an RGBPTN cell in an RGB color space in the first embodiment provided by the present invention
  • Figure 12 is a schematic diagram showing the six faces of the RGBPTN division in the first embodiment of the present invention.
  • FIG. 13 is a schematic diagram showing the relationship between an IIUOP set and an RGBUC in the first embodiment provided by the present invention.
  • Figure 14 is a structural view showing a second embodiment of the present invention.
  • Figure 15 is a structural view showing a third embodiment of the present invention. Detailed ways
  • VCS Virtual Color Space
  • the VCS may include a YUV color space, and a deformation of the YUV color space, such as a YCC/YCbCr space. Other possible virtual color spaces may also be included.
  • the VCS is also a three-dimensional space. As shown in Figure 6, the three-dimensional axes are denoted as VC1, VC2, VC3, respectively, and the sitting marks of any point p in the space (vcl, vc2, vc3)
  • Vcl 0 sr(vcl 0 , vc2 0 , vc3 0 )
  • the color positive transformation function T(.) and the inverse color transformation function S(.) can be any nonlinear function, but satisfy a set of loose conditions, and the form cannot be controlled.
  • the correction for the component vcl vc2 vc3 involves the three components of rgb, as shown in equation [6], formula [7], and formula [8], respectively: t!(g r (r r ), g g (g r ), g b (b r )) [6]
  • t 2 t 3 represents a color space transformation function, respectively.
  • c uVCS represents the VCS color vector before the gamma correction, and u represents the uncorrected.
  • c eVCS represents the corrected VCS color vector and c represents corrected (corrected).
  • the color positive transformation function T(.) from RGB to VCS and the inverse color transformation function S(.) from VCS to RGB can be arbitrary nonlinear functions, only satisfy a set of loose conditions, and The specific form cannot be controlled. Therefore, the RGB unit cube RGBUC, after mapping to the VCS color space, will be a closed single-connected set of arbitrary shapes. Therefore, if you want to perform Gamma correction in the VCS color space, you must solve the following two problems:
  • VAPS VCS admissible point set
  • the present invention has the concept of "Bounding Bounding Cuboid (MBC)".
  • MBC Bounding Bounding Cuboid
  • the smallest inclusion cuboid is the smallest of all cuboids containing VAPS.
  • the VAPS and its smallest containing cuboid are shown in Figure 7.
  • the abbreviation MBC is used below to indicate that the smallest contains a cuboid.
  • the present invention provides a first embodiment, namely a gamma school for video communication
  • a first embodiment namely a gamma school for video communication
  • the main idea is: First, the MBC of the VAPS contained in the VCS space is divided into a plurality of partitions; in each of the partitions, the optimal linear gamma correction function is determined. (The principle of determination is to be based on some optimal criterion, which needs to be associated with the Gamma property function G(.) in RGB space); then the uncorrected VCS is based on the optimal linear gamma correction function. Each sample of the signal is corrected.
  • the first part is a process of dividing the MBC of the VAPS contained in the VCS space to form a plurality of partitions and determining an optimal Gamma correction function on the partition;
  • the second part is the process of correcting each sample value of the uncorrected VCS signal based on the determined Gamma correction function.
  • the specific implementation process of the first embodiment is as shown in FIG. 8, and includes the following contents:
  • Step S201 determining the equations of the six faces of the VAPS according to the mathematical equations of the six faces of the RGB unit cube RGBUC obtained by normalizing the RGB color space and the inverse color transformation function.
  • the six faces of the VAPS are six spatial curved surfaces surrounding the VAPS.
  • the six faces: ABCD, ABFE, BCGF, CDHG, DAEH, and EFGH are mapped to the six "faces" of the VAPS by the inverse color transformation function S(.), respectively (part of the three-dimensional space surface, respectively).
  • Step S202 determining a spatial extent of the VAPS according to an equation of the six faces of the VAPS, and then determining a spatial extent of the smallest including the rectangular parallelepiped MBC according to the spatial extent of the VAPS.
  • step S202 the spatial range of the VAPS is first determined according to the equations of the six faces of the VAPS, and then the maximum coordinate and the minimum in the directions of the three different coordinate axes vcl, vc2, and vc3 are determined according to the spatial range of the VAPS. Coordinates, and then determining, according to the maximum coordinate and the minimum coordinate of the VAPS in the direction of the three different coordinate axes vcl, vc2, vc3, the maximum including the rectangular body MBC of the VAPS in the corresponding three different coordinate axes Coordinates and minimum coordinates; determining the spatial extent of the MBC according to the determined maximum and minimum coordinates of the MBC in the corresponding three different coordinate axes.
  • a maximum coordinate and a minimum coordinate of the minimum containing cuboid MBC in three different coordinate axes are equivalent to determining the minimum
  • the coordinates of the lower left rear vertex S of the cuboid MBC: (minvc 1 , minvc2, minvc3) , and the coordinates of the upper right front vertex U: (maxvc 1 , maxvc2, maxvc3) are equivalent to determining the minimum
  • minvcl represents the minimum value of the VC1 coordinates of all points in the VAPS
  • minvc2 represents the minimum value of VC2 coordinates of all points in the VAPS
  • minvc3 represents the minimum value of VC3 coordinates of all points in the VAPS
  • maxvc 1 represents the maximum value of VC1 coordinates of all points in the VAPS
  • maxvc2 represents the maximum value of VC2 coordinates of all points in the VAPS
  • maxvc3 represents the maximum value of VC3 coordinates of all points in the VAPS.
  • the range of the MBC is completely determined.
  • a maximum coordinate and a minimum coordinate of the minimum containing cuboid MBC in three different coordinate axes are equivalent to determining the minimum Contains the coordinates of the lower right vertices of the cuboid MBC, and the coordinates of the upper left vertices.
  • the coordinates of other vertices of the MBC are determined accordingly. Once the coordinates of all the vertices of the MBC are determined, the range of the MBC is completely determined.
  • the MBC divided into a plurality of rectangular sub-cells, or even in a non-uniform VC1 direction into segments; VC2 uniform in the direction or N 2 into a non-uniform section; uniform or non-uniform in the direction VC3 N 3 is divided into sections. Therefore, the MBC is divided into N T NiN 2 N 3 divisions.
  • the specific division rules are as follows:
  • the interval [minvc3, maxvc3] is uniformly or non-uniformly divided into N 3 subintervals in the VC3 direction.
  • VCSPTN is an abbreviation of VCS ParTitioN
  • VCSPTN is an abbreviation of VCS ParTitioN
  • VCSPT (i, j, k)
  • Step S204 classify all VCSPTN cells according to the relationship between each VCSPTN cell and VAPS formed by MBC segmentation, and combine the cells with non-empty intersections of VAPS to calculate the UOP (Union of Partition). .
  • Each VCSPTN cell formed by MBC segmentation has three forms of relationship with VAPS. According to the relationship between the MBC cell and VAPS, all VCSPTN cells of MBC are divided into three categories, as follows:
  • VCSPT (i, j, k) VAPS.
  • Vc3 dp 3 (k)
  • Vc2 dp 2 (j + l)
  • Vcl dp'(i)
  • the cell corresponding to the first type of VCSPTN cell is called the first class RGBPTN cell; the cell corresponding to the second class VCSPTN cell is called the second class RGBPTN cell.
  • RGBPTN(i, j, k) is not completely contained in RGBUC, but has an intersection with RGBUC (the intersection is not an empty set), ie RGBPTN(i,j,k) RGBUC ; RGBPTN(i,j,k) n RGBUC ⁇ ⁇ .
  • each RGBPTN cell there are also six faces, the labels of which are shown in Fig. 12. Each face of the RGBPTN cell corresponds to each face of the VCSPTN cell.
  • the RGBPTN division if the equation of its six faces is determined, the mathematical representation of the division is completely determined, and the spatial extent of the division as a set is completely determined.
  • the six-face equation of the division is 3 ⁇ 4:
  • Step S206 according to the union of all VCSPTN partitions having a non-empty intersection with the VAPS
  • UOP and the inverse color transform function, obtains the inverse of the UOP set in the RGB color space IIUOP and uses it as the union of all RGBPTN partitions with non-empty intersections with the RGBUC.
  • the first type and the second type of VCSPTN divisions correspond to the union of RGBPTN divisions in the RGB color space, that is, UOP maps to the set in the RGB color space by the inverse color transformation function S(.), this set is called IIUOP ( Inverse Image of UOP, the inverse of UOP ;). Obviously IIUOP completely contains RGBUC, as shown in Figure 13, which satisfies the relationship shown in equation [29]:
  • Step S207 in the IIUOP set, determining each RGBPTN cell and the RGBUC Intersection, and constructing a linearized parameter representation of the original gamma characteristic function on the intersection to obtain a corresponding linear gamma characteristic function; and determining parameters of the linear gamma characteristic function according to a set principle,
  • the principle is that the generalized average distance between the distorted RGB color vector caused by the original gamma characteristic function and the distorted RGB color vector caused by the linear gamma characteristic function reaches the set target value.
  • step S207 based on the minimum mean square error criterion, a mathematical optimization problem including the parameter of the lattice linear Gamma characteristic function is established on the intersection of each RGBPTN division and RGBUC; then the mathematical optimization problem is solved, and the Linear Gamma property function parameters.
  • the Gamma property function G(.) is defined on the whole RGBUC, it is a nonlinear function, but it can be linearized to approximate each RGBPTN partition, that is, a linear function is used in the RGBPTN division. Upwardly replaces the original Gamma property function.
  • this approximation means ⁇ under the mouth:
  • CdRGB K(i,j,k)c rRGB + B(i,j,k)
  • K(i,j,k) is a 3 3 matrix
  • B(i,j,k) is a 3 1 column vector
  • K(i, j, k) and B(i, j, k) in the Gamma property function are determined according to the following optimal criteria.
  • ⁇ dc rRGB takes the minimum value.
  • crRGB e RGBPTN(i,j ,k) RGB RGBUC ⁇ ( (i, j, k)c rRGB + B(i, j, k) - G(c rRGB )) T (K(i, j, k) c rRGB + B(i, j, k) - G(c rRGB )).
  • Step S208 determining an intersection of the VCSPTN cell corresponding to the RGBPTN cell and the VAPS, and determining a location according to a linear gamma characteristic function on an intersection of each RGBPTN and the RGBUC, and a setting principle.
  • a linear gamma correction function on the intersection; the setting principle is: the linear correction function is such that the RGB of the virtual color space color vector on the intersection of the corrected VCSPTN cell and the VAPS is inversely transformed by color The color vector, the vector-averaged average distance between the vectors and the original RGB color vector of the virtual color space color vector before the correction by the positive color conversion reaches the set target value.
  • a linear gamma correction function parameter containing the division and a corresponding RGBPTN division are established on each VCSPTN division mainly according to a minimum mean square error criterion.
  • a mathematical optimization problem of the linear gamma characteristic function parameter and then solving the mathematical optimization problem, determining the linear gamma correction function parameters, thereby determining linear gamma correction function parameters on all VCSPTN divisions, and then according to the linear gamma correction function
  • the parameters determine the linear gamma correction function on all VCSPTN divisions.
  • each VCSPTN partition the same linear function is used to represent the gamma characteristic function; in different VCSPTN divisions, the linear function It is also different.
  • a segmentation line function is used to represent the Gamma correction function in the VCSPTN division.
  • the gamma correction function (GC stands for Gamma correction) is expressed by GC(.).
  • the specific form is shown in formula [35]:
  • g Cl (.), Gc 2 (.), Gc 3 (.) Represent the generation of corrected VCS vector c cVCS vcl, vc2, vc3 component.
  • c cVCS P(i, j, k)T(K(i, j, k)c rRGB +B(i,j,k)) + Q(i,j,k) [39]
  • the P(i, j, k) parameter and the Q(, j, k) parameter in the Gamma correction function are determined below by using the VCS color vector c cVCS and the VCS in the absence of Gamma distortion.
  • the magnitude of the error of the color vector, and the principle of the Gamma correction function is determined according to the minimum error, and the P(i, j, k) parameter and the Q(, j, k) parameter are determined.
  • MS£" I /i ,, 3 ( c cvcs - T(c rRGB )) (c cVCS - T(c rRGB ))dc rRGB [41] where MSE represents the mean square error and English is Mean Square Error.
  • each RGB cell RGBPTN(i,j,k) corresponds to each cell VCSPTN(i,j,k) in the UOP set in the VCS space range, then it is defined in RGBPTN(i,j,k)
  • the Gamma property function on the ) is defined by the Gamma correction function of VCSPTN(i, j, k).
  • Other gamma gamma functions and gamma correction functions do not affect RGBPTN(i,j,k) and VCSPTN(i,j,k).
  • the Gamma correction function for each VCSPTN(i,j,k) bin can be determined independently, depending only on the Gamma characteristic function on the RGBPTN(i,j,k) bin, or more strictly depends on RGBPTN(i,j,k) A linear representation of the Gamma property function on the bin.
  • the second type of division it is considered that the second type of division is not completely included in the VAPS set, and the Gamma correction function should be defined on the VAPS set. Therefore, for each cell, the first thing to do is to determine the intersection with the VAPS set.
  • the reason for selecting the integral set as VCSPTN(i, j, k) n VAPS is that for the first type of VCSPTN partition, it is completely included in the VAPS set.
  • VCSPTN(i, j,k) n VAPS VCSPTN(i , j, k) , that is, the integral set is exactly VCSPTN (I, j, k); for the second type of VCSPTN, not included in the VAPS set, but there is a non-empty intersection, this time the set should be VCSPTN ( The intersection of I, j, k) and VAPS VCSPTN(i, j, k) n VAPS.
  • the integration set is VCSPTN(i, j, k) n VAPS for both the first type and the second type of VCSPTN.
  • the Gamma correction function is determined on a case-by-case basis. Since all the bins constitute the entire UOP set, the Gamma correction function on all UOP sets can be determined by determining the Gamma correction function on each VCSPTN(i, j , k) bin.
  • the criterion by which the Gamma correction function on each bin is determined is the minimum mean square error criterion defined by equation [34]. According to this principle, the 3 x 3 matrix of the P(i, j, k) parameter is determined, as in the formula [ 42] shows:
  • the video information sent from the transmitting end to the receiving end is actually a sequence of sample values formed by some color signal samples.
  • a sequence of images is included, and each frame of image contains a plurality of pixels, each pixel corresponding to a sample of a color vector signal. So in the end, the video can be represented as a sample of the color vector signal of a sequence.
  • each raw RGB vector electrical signal sample value included in the video information is obtained and converted into a virtual color space vector electrical signal in the virtual color space, and the corresponding VCS virtual color space vector power is obtained.
  • the signal is sampled, and then the gamma correction function of each virtual color space vector electrical signal is gamma corrected by using the gamma correction function on each VCSPTN division determined by the above process, as follows:
  • Step S209 in the video communication process, acquiring each original RGB vector electrical signal sample value included in the video information, and converting the virtual color space vector signal into a virtual color space vector signal to obtain a corresponding VCS virtual color space.
  • the vector signal is sampled, and then the VCSPTN cell to which the VCS virtual color space vector signal sample value belongs is determined.
  • step S209 the components corresponding to the virtual color space vector signal samples on three different coordinate axes are compared with the coordinate ranges of each VCSPTN cell, if: Dp 1 (i 0 ) ⁇ vcl ⁇ dp 1 (i 0 +l)
  • the VCS color vector belongs to the VCSPTN (i Q , j Q , k Q ) cell. That is, when the components of the virtual color space vector signal sample values on three different coordinate axes are within a coordinate range of a certain VCSPTN cell on the corresponding coordinate axis, the virtual color space is determined.
  • the vector signal sample value is attributed to the VCSPTN cell.
  • Step S210 correcting the VCS vector signal sample value according to the determined linear gamma correction function on the basis of the VCSPTN (i Q , j Q , k Q ) (as shown in the formula [47]).
  • Step S209 and step S210 are repeated for the sample value of the next uncorrected VCS signal in the video information until all the sample values have been processed.
  • the current standard correction method can be used to make the final signal input-output relationship linear.
  • the second embodiment provided by the present invention is a gamma correction system for video communication, and the structure thereof is as shown in FIG. 14, and includes: a cell determination unit, a linear gamma correction function determination unit, a cell matching unit, and a gamma Ma correction unit.
  • the cell determining unit includes a first spatial range determining subunit, a second spatial range determining subunit, and a cell determining subunit; the second spatial range determining subunit includes a VAPS coordinate determining subunit, and the MBC coordinate is determined. Subunits and MBC spatial extent determination subunits.
  • the linear gamma correction function determining unit includes a linear gamma characteristic function determining subunit and a linear gamma correction function determining subunit; wherein the linear gamma characteristic function determining subunit includes an RGBPTN division set determining module and linearity a gamma characteristic function determining module; wherein the cell matching unit includes a comparison subunit and a cell assignment determining subunit.
  • the division determining unit divides the VAPS according to the minimum inclusion box MBC of the virtual color space allowable point set VAPS in the virtual color space to form a plurality of VCSPTN points.
  • the specific processing process is as follows:
  • the first spatial range determining subunit determines an equation of six faces of the VAPS according to a mathematical equation of six faces of the RGB unit cube RGBUC obtained by normalizing the RGB color space, and an inverse color transformation function;
  • the six faces of the VAPS are six spatial curved surfaces surrounding the VAPS; and, according to the equations of the six faces of the VAPS, the spatial extent of the VAPS is determined; the specific implementation process is similar to the related description in the first embodiment. , will not be described in detail here.
  • the second spatial range determining subunit determines, according to the spatial extent of the VAPS determined by the first spatial range determining subunit, a spatial extent of the smallest including the rectangular parallelepiped MBC; in the specific processing, the subunit is determined by the VAPS coordinates, according to the The spatial extent of the VAPS determines its maximum coordinate and minimum coordinate in three different coordinate axis directions; and, by the MBC coordinate determining subunit, the VAPS determined by the VAPS coordinate determining subunit is in the direction of the three different coordinate axes
  • the maximum coordinate and the minimum coordinate on the upper limit determine the maximum coordinate and the minimum coordinate of the smallest BCPS containing the VAPS in the corresponding three different coordinate axes; and, by the MBC spatial range, the maximum coordinate and the minimum in the direction of the coordinate axis
  • the coordinates determine the spatial extent of the MBC.
  • the specific implementation process is similar to the related description in the first embodiment, and will not be described in detail herein.
  • the cell determining sub-unit divides the MBC determined by the second spatial range determining sub-unit into a plurality of VCSPTN cells to obtain a plurality of VCSPTN cells of the VAPS.
  • the specific implementation process is the same as the related description in the first embodiment, and will not be described in detail herein.
  • a linear gamma correction function on each VCSPTN partition the linear gamma correction function causes the virtual color space color vector on the corrected VCSPTN division to pass the inverse RGB color vector, and the color positive transformation
  • the generalized average distance between the vectors between the original RGB color vectors of the virtual color space color vector before the correction is obtained reaches the set target value.
  • the processing is performed by the linear gamma characteristic function determining subunit and the linear gamma correction function determining subunit: the linear gamma characteristic function determining subunit determines the modulo through the RGBPTN partition set therein Blocking, determining a spatial range of the corresponding inverse image bin RGBPTN in the RGB color space according to a mathematical equation of six faces of the VCSPTN cell and a color positive transform function; and, according to the non-empty intersection with the VAPS
  • the union UOP of all VCSPTN divisions, and the inverse color transformation function obtains the inverse image IIUOP of the UOP set in the RGB color space, and uses it as the union of all the RGBPTN divisions with the non-empty intersection of the RGBUC; In the IIUOP set, the intersection of each RGBPTN cell and the RGBUC is determined; the specific implementation process is the same as that in the first embodiment, and will not be described in detail herein.
  • the linear gamma characteristic function determining module constructs a linearized parameter representation of the original gamma characteristic function on the intersection determined by the RGBPTN cell set determination module, and obtains a corresponding linear gamma characteristic function;
  • the principle of determining the parameters of the linear gamma characteristic function is: a distortion RGB color vector caused by the original gamma characteristic function, and a distortion RGB caused by the linear gamma characteristic function.
  • the generalized average distance between the vectors between the color vectors reaches the set target value; the generalized average distance between the vectors refers to the mean square error.
  • the set target value is a minimum value of the generalized average distance between the vectors.
  • the linear gamma correction function determining subunit determines an intersection of a VCSPTN cell corresponding to the RGBPTN cell and the VAPS, and determines each RGBPTN and the RGBUC constructed by the subunit according to the linear gamma characteristic function.
  • the generalized average distance between the vectors refers to the mean square error.
  • the set target value is a minimum value of the generalized average distance between the vectors.
  • the cell matching unit acquires each original RGB vector electrical signal included in the video information ⁇
  • the virtual color space vector signal corresponding to the sample value is sampled, and the virtual color space vector signal sample value is corresponding to the component on three different coordinate axes by comparing the subunits, and the coordinate range of each VCSPTN cell
  • the comparison is performed, and the comparison result is transmitted to the division attribution determining subunit; the specific implementation process is the same as the related description in the first embodiment, and will not be described in detail herein.
  • the divisional belonging determination subunit determines that the comparison result is that the virtual color space vector signal sample values are corresponding to the components on the three different coordinate axes when the coordinates of a certain VCSPTN are on the corresponding coordinate axes And determining that the virtual color space vector signal sample value belongs to the VCSPTN cell.
  • the comparison method of the coordinates is the same as the related description in the first embodiment, and will not be described in detail here.
  • the gamma correction unit calls a linear gamma correction function on the VCSPTN division matched by the division matching unit, and performs gamma correction on each of the obtained virtual color space vector signal samples.
  • the specific implementation process is the same as the related description in the first embodiment, and will not be described in detail herein.
  • the third embodiment provided by the present invention is a gamma correction device whose structure is as shown in Fig. 15, and includes a division determining unit and a linear gamma correction function determining unit.
  • the cell determination unit includes a first spatial range determining subunit, a second spatial range determining subunit, and a cell determining subunit; the second spatial range determining subunit includes a VAPS coordinate determining subunit, and an MBC coordinate determining sub Unit and MBC space range determination subunits.
  • the linear gamma correction function determining unit includes a linear gamma characteristic function determining subunit and a linear gamma correction function determining subunit.
  • the linear gamma characteristic function determining subunit includes an RGBPTN cell set determining module and a linear gamma characteristic function determining module.
  • the division determining unit divides the VAPS according to the minimum inclusion box MBC of the virtual color space allowable point set VAPS in the virtual color space, and forms a plurality of VCSPTN divisions.
  • the specific processing procedure is as follows:
  • the first spatial range determining subunit determines an equation of six faces of the VAPS according to a mathematical equation of six faces of the RGB unit cube RGBUC obtained by normalizing the RGB color space, and an inverse color transformation function;
  • the six faces of the VAPS are six surrounding the VAPS The spatial surface; and, according to the equations of the six faces of the VAPS, the spatial extent of the VAPS is determined; the specific implementation process is the same as that in the first embodiment, and will not be described in detail herein.
  • the second spatial range determining subunit determines, according to the spatial extent of the VAPS determined by the first spatial range determining subunit, a spatial extent of the smallest including the rectangular parallelepiped MBC; in the specific processing, the subunit is determined by the VAPS coordinates, according to the The spatial extent of the VAPS determines its maximum coordinate and minimum coordinate in three different coordinate axis directions; and, by the MBC coordinate determining subunit, the VAPS determined by the VAPS coordinate determining subunit is in the direction of the three different coordinate axes
  • the maximum coordinate and the minimum coordinate on the upper limit determine the maximum coordinate and the minimum coordinate of the smallest BCPS containing the VAPS in the corresponding three different coordinate axes; and, by the MBC spatial range, the maximum coordinate and the minimum in the direction of the coordinate axis
  • the coordinates determine the spatial extent of the MBC.
  • the specific implementation process is similar to the related description in the first embodiment, and will not be described in detail herein.
  • the cell determining sub-unit divides the MBC determined by the second spatial range determining sub-unit into a plurality of VCSPTN cells to obtain a plurality of VCSPTN cells of the VAPS.
  • the specific implementation process is similar to the related description in the first embodiment, and will not be described in detail herein.
  • the linear gamma correction function determining unit on the partition constructs a linear gamma correction function on each VCSPTN partition obtained by the division determining unit, the linear gamma correction function making the corrected VCSPTN
  • the virtual color space color vector on the grid is the RGB color vector obtained by inverse color transformation, and the generalized average distance between the vectors between the original RGB color vectors of the virtual color space color vector before correction by the color positive transformation reaches the set target. value.
  • the processing of the subunit is determined by the linear gamma characteristic function determining subunit and the linear gamma correction function:
  • the linear gamma characteristic function determines a sub-unit through which an RGBPTN division set determination module determines a corresponding inverse image in the RGB color space according to a mathematical equation of six faces of the VCSPTN division and a color positive transformation function.
  • the linear gamma characteristic function determining module constructs a linearized parameter representation of the original gamma characteristic function on the intersection determined by the RGBPTN cell set determination module, and obtains a corresponding linear gamma characteristic function;
  • the principle of determining the parameters of the linear gamma characteristic function is: a distortion RGB color vector caused by the original gamma characteristic function, and a distortion RGB caused by the linear gamma characteristic function.
  • the generalized average distance between the vectors between the color vectors reaches the set target value; the generalized average distance between the vectors refers to the mean square error.
  • the set target value is a minimum value of the generalized average distance between the vectors.
  • the linear gamma correction function determining subunit determines an intersection of a VCSPTN cell corresponding to the RGBPTN cell and the VAPS, and determines each RGBPTN and the RGBUC constructed by the subunit according to the linear gamma characteristic function.
  • the generalized average distance between the vectors refers to the mean square error.
  • the set target value is a minimum value of the generalized average distance between the vectors.
  • the present invention first divides the VAPS according to the minimum included rectangular body MBC of the virtual color space allowable point set VAPS in the virtual color space to form a plurality of VCSPTN points. Then construct a linear gamma correction function on each VCSPTN partition; then acquire each original RGB vector telecommunications contained in the video information The virtual color space vector signal corresponding to the sample value, and determine the value to which it belongs.
  • the VCSPTN is divided into grids; and according to the linear gamma correction function on the VCSPTN grid, the gamma correction is performed on each of the obtained virtual color space vector signal samples. Therefore, according to the present invention, gamma correction can be implemented in various color spaces used in the intermediate processing link of the video communication process, thereby expanding the application range of the gamma correction, and solving the current gamma cannot be implemented in the intermediate processing of the video communication. Correction puzzle.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Image Processing (AREA)
  • Processing Of Color Television Signals (AREA)

Abstract

La présente invention concerne un procédé, un système et un dispositif de correction de gamma. Le procédé consiste à se procurer la partition VCSPTN (partition d'espace couleur virtuel) en partitionnant le cuboïde limitant minimum (Minimum Bounding Cuboid) de l'ensemble VAPS (ensemble des points admissibles de l'espace couleur virtuel) dans le VCS (espace couleur virtuel), à construite une fonction linéaire de correction gamma de la partition VCSPTN, à se procurer la valeur d'échantillonnage du signal du vecteur virtuel de couleur correspondant à la valeur d'échantillonnage du signal électrique du vecteur RGB original, et à déterminer la partition VCSPTN à laquelle elle appartient, et à la corriger sur la base de la fonction linéaire de correction de gamma de la partition VCSPTN. L'invention permet ainsi d'étendre le champ d'application de la correction de gamma, de façon à trouver une solution au fait que la correction de gamma ne peut pas s'appliquer couramment à la phase de traitement intermédiaire de communication vidéo.
PCT/CN2008/070122 2007-01-17 2008-01-17 Procédé, système et dispositif de correction de gamma WO2008089683A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN200710000989.6 2007-01-17
CN200710000989A CN100581271C (zh) 2007-01-17 2007-01-17 伽玛校正方法、伽玛校正系统和伽玛校正装置

Publications (1)

Publication Number Publication Date
WO2008089683A1 true WO2008089683A1 (fr) 2008-07-31

Family

ID=39644124

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2008/070122 WO2008089683A1 (fr) 2007-01-17 2008-01-17 Procédé, système et dispositif de correction de gamma

Country Status (2)

Country Link
CN (1) CN100581271C (fr)
WO (1) WO2008089683A1 (fr)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105185350A (zh) * 2015-09-23 2015-12-23 上海大学 一种支持伽玛校正的分形扫描显示控制系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1291403A (zh) * 1998-12-21 2001-04-11 皇家菲利浦电子有限公司 近似线性的伽玛数字校正
CN1658679A (zh) * 2004-02-16 2005-08-24 钰瀚科技股份有限公司 色彩校正的方法
US20060098024A1 (en) * 2004-10-18 2006-05-11 Makoto Kohno Digital video signal data processor
WO2006057126A1 (fr) * 2004-11-26 2006-06-01 Ryobi System Solutions Processeur de pixels

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1291403A (zh) * 1998-12-21 2001-04-11 皇家菲利浦电子有限公司 近似线性的伽玛数字校正
CN1658679A (zh) * 2004-02-16 2005-08-24 钰瀚科技股份有限公司 色彩校正的方法
US20060098024A1 (en) * 2004-10-18 2006-05-11 Makoto Kohno Digital video signal data processor
WO2006057126A1 (fr) * 2004-11-26 2006-06-01 Ryobi System Solutions Processeur de pixels

Also Published As

Publication number Publication date
CN100581271C (zh) 2010-01-13
CN101227622A (zh) 2008-07-23

Similar Documents

Publication Publication Date Title
RU2710888C2 (ru) Способ и устройство кодирования и декодирования цветной картинки
RU2710291C2 (ru) Способы и устройства для кодирования и декодирования цветного изображения hdr
CN107203974A (zh) 扩展的高动态范围hdr到hdr色调映射的方法、装置和系统
WO2016154336A1 (fr) Sous-échantillonnage de la chrominance et remise en forme de gamut
RU2710873C2 (ru) Способ и устройство для декодирования цветного изображения
EP3251336A1 (fr) Procédé et dispositif de mise en concordance de couleurs entre images en couleurs de dynamiques différentes
WO2023010754A1 (fr) Procédé et appareil de traitement d'image, équipement terminal et support d'enregistrement
CN110785989A (zh) 用于色域映射的方法和设备
EP2213087A1 (fr) Procédé servant à décrire une gamme de couleurs actuelle d'un dispositif de couleur ou d'un contenu de couleur
CN109845262B (zh) 用于将高动态范围画面编码/解码为编码比特流的方法和装置
CN100551080C (zh) 视频通讯中视频输入设备伽玛特性校正方法和装置
TW200521955A (en) Gamut conversion system and methods
EP2684158A1 (fr) Procédé de représentation et de traitement efficaces de données de pixel couleur dans images de pathologie numériques
EP3255874A1 (fr) Procédé de mappage de couleurs d'une image source utilisant une lut ayant des couleurs d'entrée à l'extérieur d'une gamme de couleurs source
EP3453177A1 (fr) Procédé et appareil de codage/décodage d'un nombre entier scalaire en un paramètre représentatif d'un point de pivot d'une fonction linéaire par morceaux
EP3590252B1 (fr) Mappage de gamme de couleurs de changement de teinte
WO2008089683A1 (fr) Procédé, système et dispositif de correction de gamma
CN102752622B (zh) 一种色彩空间三维显示方法
WO2018114509A1 (fr) Procédé de mappage de gamme de couleurs pour des couleurs d'entrée de contenu ldr d'entrée sur des couleurs de sortie formant du contenu hdr de sortie
CN114788280A (zh) 视频的编解码方法与装置
CN111434113A (zh) 用于高动态范围重建的饱和控制
JP2002335536A (ja) 輝度調整装置およびその動作制御方法
CN100563347C (zh) 视频通讯中视频显示设备伽玛特性校正方法和装置
Vandenberg et al. A survey on 3d-lut performance in 10-bit and 12-bit hdr bt. 2100 pq
CN101924949B (zh) 影像处理方法与影像处理系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 08700779

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 08700779

Country of ref document: EP

Kind code of ref document: A1