CN101227622A - Method , system and device for gamma correction - Google Patents

Abstract

The invention discloses a gamma correction method, a gamma correction system and a gamma correction device, the gamma correction method comprises the following steps: segmenting a minimum inclusion rectangular parallelepiped of an aggregation VAPS in virtual color space, constructing the linear gamma correction function on a VCSPTN frame, obtaining a sample value of the virtual color space vector signal which is corresponded with the sample value of the original RGB vector electrical signal, and ascertaining the VCSPTN frame where the sample value of the original RGB vector electrical signal belongs to, and correcting according to the linear gamma correction function of the VCSPTN frame. The invention can extend the applied scope of the gamma correction, which solves the difficult problem that gamma correction can not be implemented on the middle disposal tache of video communication at present.

Description

Gamma revision method, Gamma correction system and Gamma correction device

Technical field

The present invention relates to the communications field, relate in particular to the Gamma correction technology that is used for video communication.

Background technology

Along with developing rapidly of broadband network,, be called for short video communication service and also be widely used day by day based on the business of video communication.For example, video conference and visual telephone service are becoming the basic service on the NGN (NextGeneration Network, next generation network).

The video communication system framework of realization video communication service comprises video input apparatus, first color transform module, first signal processing module, network, secondary signal processing module, second color transform module and picture output device as shown in Figure 1.Its operation principle is as follows: vision signal is caught by video input apparatus as the rgb light signal, form the RGB signal of telecommunication, carry out the color direct transform by described first color transform module then, the described RGB signal of telecommunication is changed into the YUV signal of telecommunication, by described first signal processing module described YUV signal of telecommunication is carried out signal processing (compressed encoding etc.) then, carry out signal processing (removing compression coding etc.) for the secondary signal processing module by Network Transmission afterwards, and then carry out the color inverse transformation by second color transform module described YUV signal is converted into the RGB signal of telecommunication, convert light signal when finally on picture output device, showing again to.

As can be seen, the video input apparatus at two ends and video display apparatus are positioned at the RGB color space, and middle various signal processing (compressed encoding removes compression coding etc.) and network transport process carry out at the YUV color space.

Described RGB color space is by red, and is green, and blue (Red Green Blue, the RGB) linear combination of three primary colors (being that the different proportion addition mixes) generates.In the RGB color space, if every kind of color is represented its coordinate with RGB component (component), a point in can corresponding 3 dimension spaces.

Described YUV color space is introduced in order to improve signal processing efficiency in the various signal processing of centre and network transport process, it is color space commonly used in the video processing procedure (has a lot of mutation such as YCC/YCbCr space etc.), and its expression formula is shown in formula [1]:

$\left[\begin{array}{c}y\\ u\\ v\end{array}\right]=\left[\begin{array}{ccc}0.299& 0.587& 0.114\\ -0.147& -0.289& 0.436\\ 0.615& -0.515& -0.100\end{array}\right]\left[\begin{array}{c}r\\ g\\ b\end{array}\right]..............\left[1\right]$

Corresponding inverse transformation is shown in formula [2]:

$\left[\begin{array}{c}r\\ g\\ b\end{array}\right]=\left[\begin{array}{ccc}1.0000& 0& 1.1398\\ 1.0000& -0.3946& -0.5805\\ 1.0000& 2.0320& -0.0005\end{array}\right]\left[\begin{array}{c}y\\ u\\ v\end{array}\right]..............\left[2\right]$

By expression formula as can be seen color space in fact be exactly a kind of coordinate system, so the conversion between the color space comes down to a kind of coordinate system transformation (coordinate system transform), follow in the mathematics principle about coordinate system transformation.This conversion can be linear, also can be non-linear.For of the conversion of RGB color space, be linear to the YUV color space.

Conversion principle between the color space as shown in Figure 2.As seen from Figure 2, with a kind of color, if use the RGB representation in components, its component is respectively r, g, b, these components be exactly this color as the coordinate of a point in the RGB color space, and if transform in the YUV color space, corresponding coordinate will be different.

Because in each processing links of vision signal process, concrete brightness or carrier chrominance signal grade can be different, such as 256,64 grades etc., but can change in [1,1] interval, therefore by brightness or carrier chrominance signal highest ranking divided by each processing links, in signal processing, usually color space is adopted normalization (normalized) expression.For the RGB color space because r, g, b all get on the occasion of, therefore adopt normalization to represent after, require 0≤r, g, b≤1, i.e. r, g, the absolute value of b component is not more than 1; In the YUV color space, require-1≤y, u, v≤1.

In the RGB color space, if carried out the normalization processing, the RGB color space is exactly a unit cube [0,1] * [0,1] * [0,1], perhaps writes a Chinese character in simplified form work [0,1] ³, wherein multiplication sign " * " is represented the cartesian product (Cartesian Product) of set.Therefore this normalization cube that obtains is called RGB unit cube (RGB Unit Cube is called for short RGBUC).As shown in Figure 3.

When realizing video communication service based on above-mentioned video communication system framework as shown in Figure 1, influence QoS (the Quality of Service of the factor of end user's experience except network, service quality) parameter (comprises packet loss, postpone, shake, parameters such as the R factor) outside, also have because the caused distortion of Gamma characteristic (Distortion) factor of each link for luminance signal.For the former, need to guarantee the qos parameter and the pre-process and post-process (Pre-processing relevant of network with video compression coding, post-processing), for the brightness distortion problem that causes owing to the Gamma characteristic issues, need carry out gamma (Gamma) treatment for correcting, so that the final input/output relation of signal is a linear relationship.Video input apparatus (for example video camera/camera) reaches high-quality display effect when video captured/rest image shows on display device like this, obtains good user experience.

Described Gamma characteristic, it is a kind of nonlinear relation that the luminance signal that is meant certain link is imported an output relation.Through the luminance signal after distorting after the Gamma nonlinear element, it increases progressively according to the power function rule.

In the RGB color space, R, G, the B three primary colors each have Gamma characteristic separately.Be equivalent to:

r _d＝g _r(r _r)；

g _d＝g _g(g _r)；

b _d＝g _b(b _r)；.............[6]

Wherein, g _r, g _g, g _bRepresent R respectively, B, G component Gamma characteristic separately.R wherein _rRepresent original R component signal respectively, subscript r=raw represents original; r _dThe R component signal of expression process Gamma distortion, subscript d=distorted, the expression distortion.G is arranged as a same reason _d, g _r, b _d, b _rFig. 4 has provided the Gamma characteristic curve of a video input apparatus, and is wherein red, green, and blue curve is representative function g respectively _r, g _g, g _bCorresponding Gamma characteristic curve.

The prior art relevant with the present invention provided by built-in Gamma correction module in the high-end video input apparatus of part, based on the RGB color space, the signal that captures is carried out the operation principle that Gamma proofreaies and correct, as shown in Figure 5, add a Gamma correction module at correction link the luminance signal of input is carried out the Gamma correction, described Gamma property list is shown Gc (.), can proofread and correct the Gamma characteristic by described Gamma correction module is the distortion that the Gamma characteristic of Gc (.) causes, and makes that final input/output relation is a linear relationship.

Prior art has only provided to use in the RGB color space scope carries out the working condition that Gamma proofreaies and correct, the problem of dtmf distortion DTMF that the Gamma characteristic issues causes but the intermediate treatment link in the video communication process still exists, and, the color space that uses in the intermediate treatment link not only is confined to the YUV color space, be possible the color space of other kind, therefore implement Gamma correction in the shades of colour space of the intermediate treatment link in the video communication process, also exist a difficult problem.

Summary of the invention

Embodiments of the invention provide a kind of gamma revision method, Gamma correction system and Gamma correction device, by the present invention, can implement Gamma correction in the shades of colour space that the intermediate treatment link in the video communication process is used.

Embodiments of the invention are realized by the following technical solutions:

Embodiments of the invention provide a kind of gamma revision method that is used for video communication, and it comprises:

Allow the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space in the virtual color space, VAPS is divided format, form a plurality of VCSPTN lattices;

Construct the linear Gamma correction function on each VCSPTN lattice;

Obtain the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that comprises in the video information, and determine its VCSPTN lattice that is belonged to;

According to the linear Gamma correction function on the described VCSPTN lattice, resulting each virtual color space vector signal sampled value is carried out Gamma correction.

Embodiments of the invention also provide a kind of Gamma correction system that is used for video communication, and it comprises:

The lattice determining unit is used for allowing the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space of virtual color space, VAPS is divided format, and forms a plurality of VCSPTN lattices;

Linear Gamma correction function determining unit is used to construct the linear Gamma correction function on each VCSPTN lattice;

The lattice matching unit is used for obtaining the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that video information comprises, and determines its VCSPTN lattice that is belonged to;

Gammate is used for according to the linear Gamma correction function on the described VCSPTN lattice, and resulting each virtual color space vector signal sampled value is carried out Gamma correction.

Embodiments of the invention also provide a kind of Gamma correction device, and it comprises:

The lattice determining unit is used for allowing the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space of virtual color space, VAPS is divided format, and forms a plurality of VCSPTN lattices;

Linear Gamma correction function determining unit is used to construct the linear Gamma correction function on each VCSPTN lattice.

The specific embodiments that is provided by the embodiment of the invention described above as can be seen, the present invention at first allows the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space in the virtual color space, VAPS divided format, form a plurality of VCSPTN lattices; Construct the linear Gamma correction function on each VCSPTN lattice then; Then obtain the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that comprises in the video information, and determine its VCSPTN lattice that is belonged to; And, resulting each virtual color space vector signal sampled value is carried out Gamma correction according to the linear Gamma correction function on the described VCSPTN lattice.Therefore by the present invention, can in the color space that the intermediate treatment link in the video communication process is used, implement Gamma correction, thereby the range of application of Gamma correction is enlarged, solved the current difficult problem that on video communication intermediate treatment link, can't implement Gamma correction.

Description of drawings

The video communication system Organization Chart that Fig. 1 provides for background technology;

The conversion principle signal that Fig. 2 provides for background technology from the RGB color notation conversion space to the YUV color space;

After the RGB color space that Fig. 3 provides for background technology carries out the normalization processing, the RGB unit cube that obtains;

The Gamma performance diagram of the video input apparatus that Fig. 4 provides for background technology;

The correction principle schematic diagram of the Gamma distortion that the video input apparatus that Fig. 5 provides for background technology is introduced;

Fig. 6 is the schematic diagram of the VCS color space of the present invention's introducing;

Fig. 7 is VAPS among the VCS of the present invention's introducing and the minimum schematic diagram that concerns that comprises cuboid thereof;

Fig. 8 is the flow chart of first embodiment provided by the invention;

Fig. 9 is among first embodiment provided by the invention, to the principle schematic of lattice VAPS set;

Figure 10 is among first embodiment provided by the invention, the label schematic diagram of six faces of VCSPTN lattice;

Figure 11 is among first embodiment provided by the invention, and the VCSPTN lattice is with the corresponding relation figure of RGBPTN lattice in the RGB color space;

Figure 12 is among first embodiment provided by the invention, the label schematic diagram of six faces of RGBPTN lattice;

Figure 13 is among first embodiment provided by the invention, and IIUOP set and RGBUC concern schematic diagram;

Figure 14 is the structure chart of second embodiment provided by the invention;

Figure 15 is the structure chart of the 3rd embodiment provided by the invention.

Embodiment

The employed shades of colour of intermediate treatment link space in the video communication process for convenience of description, embodiments of the invention are introduced virtual color space (Virtual Color Space, notion VCS).Described VCS can comprise the YUV color space, and the distortion of described YUV color space, as YCC/YCbCr space etc.Also can comprise the virtual color space that other is possible.

Therefore consider that colour switching does not generally increase the dimension in space, described VCS still be 3 dimension spaces, and as shown in Figure 6, its 3-D walls and floor is remembered respectively and made VC1, VC2, and VC3, and the coordinate of any 1 p note work in the space (vc1, vc2, vc3).

Based on present transformation idea from the RGB color to the YUV color, the R shown in formula [1] is satisfied in the mathematic(al) manipulation of employed color direct transform function in the time of can thinking from the RGB colour switching to the VCS color ³→ R ³Function T:

$\left[\begin{array}{c}\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1\\ \mathrm{vc}2\\ \mathrm{vc}3\end{array}\right]=T\left(\left[\begin{array}{c}r\\ g\\ b\end{array}\right]\right)=\left[\begin{array}{c}{t}_1(r,g,b)\\ t_2(r,g,b)\\ t_3(r,g,b)\end{array}\right]...............\left[1\right]$

For transforming function transformation function T (.), satisfy following condition:

(1), continuous everywhere, smooth everywhere leading is promptly for any 1 p of rgb space ₀(r ₀, g ₀, b ₀), following partial derivative

Exist, that is to say that Jacobi matrix (Jacobian) exists, shown in formula [2]:

$J(r_0,g_0,b_0)={\left[\begin{array}{ccc}\frac{\&PartialD;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}{\&PartialD;r}& \frac{\&PartialD;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}{\&PartialD;g}& \frac{\&PartialD;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}{\&PartialD;b}\\ \frac{\&PartialD;\mathrm{vc}2}{\&PartialD;r}& \frac{\&PartialD;\mathrm{vc}2}{\&PartialD;g}& \frac{\&PartialD;\mathrm{vc}2}{\&PartialD;b}\\ \frac{\&PartialD;\mathrm{vc}3}{\&PartialD;r}& \frac{\&PartialD;\mathrm{vc}3}{\&PartialD;g}& \frac{\&PartialD;\mathrm{vc}3}{\&PartialD;b}\end{array}\right]}_{\begin{array}{c}r=r0\\ g=g0\\ b=b0\end{array}}..................\left[2\right]$

(2), this conversion is reversible everywhere, for any 1 p of rgb space ₀(r ₀, g ₀, b ₀), if q ₀=T (p ₀), q ₀Be among the VCS a bit, have color counter-transformation function S so, make p ₀=S (q ₀).Sufficient and necessary condition is: for any 1 p of rgb space ₀(r ₀, g ₀, b ₀), Jacobi matrix (JaGobian) reversible (having inverse matrix); Perhaps equivalently determinant shown in formula [3]:

det(J(r ₀，g ₀，b ₀))≠0........................................................[3]

Therefore, under above-mentioned two preconditions, there is the color counter-transformation function S (.) of the overall situation corresponding with it in color direct transform function T (.), and S (.) also satisfies this two conditions.Because T (.) and S (.) are inverse transformations each other, so for any p ₀(r ₀, g ₀, b ₀), satisfy:

$S\left(T\right(\left[\begin{array}{c}{r}_0\\ g_0\\ b_0\end{array}\right]\left)\right)=S\left(\left[\begin{array}{c}{t}_1(r_0,g_0,b_0)\\ t_2(r_0,g_0,b_0)\\ t_3(r_0,g_0,b_0)\end{array}\right]\right)$

$=\left[\begin{array}{c}\mathrm{sr}(t_1(r_0,g_0,b_0),t_2(r_0,g_0,b_0),t_3(r_0,g_0,b_0))\\ \mathrm{sg}(t_1(r_0,g_0,b_0),t_2(r_0,g_0,b_0),t_3(r_0,g_0,b_0))\\ \mathrm{sb}(t_1(r_0,g_0,b_0),t_2(r_0,g_0,b_0),t_3(r_0,g_0,b_0))\end{array}\right]=\left[\begin{array}{c}{r}_0\\ g_0\\ b_0\end{array}\right]..............\left[4\right]$

For any q ₀(vc1 ₀, vc2 ₀, vc3 ₀), satisfy the relation shown in formula [5]:

$T\left(S\right(\left[\begin{array}{c}{\mathrm{vc}1r}_0\\ {\mathrm{vc}2}_0\\ {\mathrm{vc}3}_0\end{array}\right]\left)\right)=\left[\begin{array}{c}\mathrm{sr}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0)\\ \mathrm{sg}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0)\\ \mathrm{sb}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0)\end{array}\right]$

$=\left[\begin{array}{c}{t}_1(\mathrm{sr}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0),\mathrm{sg}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0),\mathrm{sb}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0))\\ t_2(\mathrm{sr}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0),\mathrm{sg}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0),\mathrm{sb}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0))\\ t_3(\mathrm{sr}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0),\mathrm{sg}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0),\mathrm{sb}({\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0,{\mathrm{vc}2}_0,{\mathrm{vc}3}_0))\end{array}\right]$

$=\left[\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1}_0\\ {\mathrm{vc}2}_0\\ {\mathrm{vc}3}_0\end{array}\right]...............\left[5\right]$

Color direct transform function T (.) and color counter-transformation function S (.) can be any nonlinear functions, just satisfy one group of loose condition, and form is uncontrollable.

Carry out in the VCS space if Gamma proofreaies and correct, so for component vc1, vc2, vc3 proofread and correct and relate to r, g, and three components of b, respectively shown in formula [6], formula [7] and formula [8]:

vc1＝t ₁(r _d，g _d，b _d)＝t ₁(g _r(r _r)，g _g(g _r)，g _b(b _r))............................[6]

vc2＝t ₂(r _d，g _d，b _d)＝t ₂(g _r(r _r)，g _g(g _r)，g _b(b _r))............................[7]

vc3＝t ₃(r _d，g _d，b _d)＝t ₃(g _r(r _r)，g _g(g _r)，g _b(b _r))............................[8]

Wherein, t ₁, t ₂, t ₃Represent the color notation conversion space function respectively.Vc1 as can be seen, vc2 intercouples between the vc3 component.If the form with matrix and vector is represented, shown in formula [9]:

c _uVCS＝T(c _dRGB)＝T(G(c _rRGB))...........................................[9]

Be defined as follows symbol:

$c_{\mathrm{uVCS}}=\left[\begin{array}{c}{\mathrm{vc}1}_u\\ {\mathrm{vc}2}_u\\ {\mathrm{vc}3}_u\end{array}\right],$ $c_{\mathrm{cVCS}}=\left[\begin{array}{c}{\mathrm{vc}1}_c\\ {\mathrm{vc}2}_c\\ {\mathrm{vc}3}_c\end{array}\right],$ $c_{\mathrm{rRGB}}=\left[\begin{array}{c}{r}_r\\ g_r\\ b_r\end{array}\right],$ $c_{\mathrm{dRGB}}=\left[\begin{array}{c}{r}_d\\ g_d\\ b_d\end{array}\right],$ $G(.)=\left[\begin{array}{c}{g}_r(.)\\ g_g(.)\\ g_b(.)\end{array}\right]$

C wherein _UVCSVCS color vector before expression is not proofreaied and correct through Gamma, u represents uncorrected (uncorrected).Accordingly, c _CVCSVCS color vector after expression is proofreaied and correct, c represents corrected (having proofreaied and correct).c _RRGBRepresent undistorted RGB color vector, r represents raw (original, undistorted); c _DRGBRGB color vector after the expression distortion, r represents distorted (distortion).

From as can be seen above-mentioned,, just satisfy one group of loose condition, and concrete form is uncontrollable because color direct transform function T (.) and the color counter-transformation function S (.) from VCS to RGB from RGB to VCS can be any nonlinear functions.Therefore RGB unit cube RGBUC, be mapped to the VCS color space after, can be that company's single-pass closed set of an arbitrary shape is closed (closed single-connected set).Therefore can expect, proofread and correct, must solve following two problems if carry out Gamma at the VCS color space:

Does 1, g, b component Gamma characteristic separately obtain directly carrying out the correction function that Gamma proofreaies and correct how from r in the VCS color space?

2, how effectively company's single-pass closed set of arbitrary shape to be closed and cut apart, form a plurality of lattices (partition), in each lattice, carry out linear Gamma then and proofread and correct?

In addition, consider in the VCS color space, be not have a few and all have meaning in practice, have only the point transformation that those can be from RGBUC and come, be only the actual physics meaning, allow, then with in the VCS color space all like this set of the point of " permission " be called VAPS (VCS admissiblepoint set, VCS allow the some set).Therefore in the VCS color space, directly carry out correction function that Gamma proofreaies and correct and be based on that described VAPS set determines.

Because the shape of lattice is simple more, when carrying out the Gamma correction, for a given VCS signal vector, judge that its deterministic process that whether is positioned at certain lattice calculating is just simple more, Gamma proofreaies and correct efficient and just can improve, but owing to the VAPS out-of-shape, can't fully just in time be divided into the lattice of a plurality of regular shapes, therefore the present invention introduces minimum and comprises cuboid (Minimum Bounding Cuboid, notion MBC).It is exactly to comprise of volume minimum in the cuboid of VAPS at all that so-called minimum comprises cuboid.VAPS and minimum thereof comprise cuboid as shown in Figure 7.Represent that with abbreviation MBC minimum comprises cuboid below.

The invention provides first embodiment based on above-mentioned consideration, promptly a kind of gamma revision method that is used for video communication, its main thought is: at first the MBC of the VAPS that comprises in the VCS space is divided and format, form a plurality of lattices (partition); In described each lattice, determine optimum linear Gamma correction function (definite principle is to carry out according to certain optiaml ciriterion, and this need carry out related with the Gamma characterisitic function G (.) in the rgb space); According to the linear Gamma correction function of described optimum, each sampled value of uncorrected VCS signal is proofreaied and correct then.

When implementing described first embodiment, comprise two parts content concrete:

First is that the MBC to the VAPS that comprises in the VCS space divides and formats, and forms a plurality of lattices (partition), and determines the process of the optimum Gamma correction function on the described lattice;

Second portion is according to determined Gamma correction function, the process that each sampled value of uncorrected VCS signal is proofreaied and correct.

The specific implementation process of described first embodiment comprises following content as shown in Figure 8:

Step S201 according to the RGB color space being carried out the math equation that six faces of resulting RGB unit cube RGBUC are handled in normalization, and the color counter-transformation function, determines the equation of six faces of VAPS.Six faces of wherein said VAPS are for surrounding six space curved surfaces of described VAPS.

Under the one group of assumed condition that satisfies about color direct transform function T (.), there is the color counter-transformation function S (.) of the overall situation corresponding with it in color direct transform function T (.), RGBUC=[0 like this, 1] ³6 face: ABCD, ABFE, BCGF, CDHG, DAEH and EFGH, be mapped to 6 " faces " (being respectively the part of 3 dimension space curved surfaces) of VAPS respectively by described color counter-transformation function S (.).

6 " faces " of described VAPS intersect to form " limit " (space curve one section), and " limit " intersected and formed summit.Topological relation between these " faces ", " limit " and the summit (topological relationship) is identical with topological relation between face, limit and the summit of RGBUC.Therefore, exist following summit to arrive " summit ", face arrive " limit " to " face " and limit mapping relations, as follows:

1, " summit " arrived on the summit:

A→A’，B→B’，C→C’，D→D’，E→E’，F→F’，G→G’，H→H’

2, face arrives " face ":

ABCD→A’B’C’D’，ABFE→A’B’F’E’，BCGF→B’C’G’F’

CDHG→C’D’H’G’，DAEH→D’A’E’H’，EFGH→E’F’G’H’

3, " limit " arrived on the limit:

AB→A’B’，BC→B’C’，CD→C’D’，DA→D’A’，EF→E’F’，FG→F’G’，GH→G’H’，HE→H’E’，AE→A’E’，BF→B’F’，CG→C’G’，DH→D’H’

For each " face " of VAPS, can represent with an equation, as follows respectively:

A ' B ' C ' D ' face:

$\left\{\begin{array}{c}\mathrm{sb}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)=0\\ 0\&le;\mathrm{sr}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\\ 0\&le;\mathrm{sg}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\end{array}\right................\left[10\right]$

A ' B ' F ' E ' face:

$\left\{\begin{array}{c}\mathrm{sr}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)=1\\ 0\&le;\mathrm{sg}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\\ 0\&le;\mathrm{sb}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\end{array}\right................\left[11\right]$

B ' C ' G ' F ' face:

$\left\{\begin{array}{c}\mathrm{sg}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)=1\\ 0\&le;\mathrm{sr}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\\ 0\&le;\mathrm{sb}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\end{array}\right................\left[12\right]$

C ' D ' H ' G ' face:

$\left\{\begin{array}{c}\mathrm{sr}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)=0\\ 0\&le;\mathrm{sg}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\\ 0\&le;\mathrm{sb}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\end{array}\right................\left[13\right]$

D ' A ' E ' H ' face:

$\left\{\begin{array}{c}\mathrm{sg}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)=0\\ 0\&le;\mathrm{sr}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\\ 0\&le;\mathrm{sb}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\end{array}\right................\left[14\right]$

E ' F ' G ' H ' face:

$\left\{\begin{array}{c}\mathrm{sb}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)=0\\ 0\&le;\mathrm{sr}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\\ 0\&le;\mathrm{sg}(\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1,\mathrm{vc}2,\mathrm{vc}3)\&le;1\end{array}\right................\left[15\right]$

According to the equation of six faces of above-mentioned VAPS, just can determine the spatial dimension of VAPS, with and at 3-D walls and floor vc1, vc2, maximum coordinates on the vc3 direction and min coordinates.

Step S202 according to the equation of six faces of described VAPS, determines the spatial dimension of VAPS, determines the spatial dimension that its minimum comprises cuboid MBC according to the spatial dimension of described VAPS then.

Among the step S202, at first according to the equation of six faces of VAPS, determine the spatial dimension of VAPS, determine that according to the spatial dimension of described VAPS it is at three different reference axis vc1 then, vc2, maximum coordinates on the vc3 direction and min coordinates, then according to described VAPS at described three different reference axis vc1, vc2, the maximum coordinates on the vc3 direction comprises maximum coordinates and the min coordinates of cuboid MBC on three different change in coordinate axis direction of correspondence with the minimum that min coordinates is determined described VAPS; According to maximum coordinates and the min coordinates spatial dimension of determining described MBC of the MBC that is determined on three different change in coordinate axis direction of correspondence.

When the virtual color space of described VCS adopts left-handed coordinate system, determine that described minimum comprises cuboid MBC at three different reference axis (VC1, VC2, VC3) maximum coordinates on the direction and min coordinates are equivalent to and determine that described minimum comprises the coordinate of summit S behind the lower-left of cuboid MBC: (minvc1, minvc2, minvc3) and the coordinate of upper right preceding summit U: (maxvc1, maxvc2, maxvc3).

Wherein minvc1 represents that a minimum value of VC1 coordinate is arranged among the VAPS;

Wherein minvc2 represents that a minimum value of VC2 coordinate is arranged among the VAPS;

Wherein minvc3 represents that a minimum value of VC3 coordinate is arranged among the VAPS;

Wherein maxvc1 represents that a maximum of VC1 coordinate is arranged among the VAPS;

Wherein maxvc2 represents that a maximum of VC2 coordinate is arranged among the VAPS;

Wherein maxvc3 represents that a maximum of VC3 coordinate is arranged among the VAPS.

The coordinate minvc1 of ordering then according to S, minvc2, the coordinate maxvc1 of minvc3 and U, maxvc2, maxvc3 correspondingly determines the coordinate on other summit of MBC, and is as follows respectively:

P(maxvc1，minvc2，minvc3)；

Q(maxvc1，maxvc2，minvc3)；

R(minvc1，maxvc2，minvc3)；

T(maxvc1，minvc2，maxvc3)；

V(minvc1，maxvc2，maxvc3)；

W(minvc1，minvc2，maxvc3)。

After having determined the coordinate on all summits of MBC, just determined the scope of MBC fully.

When the virtual color space of described VCS adopts right-handed coordinate system, determine that described minimum comprises cuboid MBC at three different reference axis (VC1, VC2, VC3) maximum coordinates on the direction and min coordinates are equivalent to and determine that described minimum comprises the coordinate on summit behind the cuboid MBC bottom right and the coordinate on upper left preceding summit.

According to the coordinate on summit behind the bottom right of being determined and the coordinate on upper left preceding summit, correspondingly determine the coordinate on other summit of MBC then.After having determined the coordinate on all summits of MBC, just determined the scope of MBC fully.

Step S203, with MBC at vc1, vc2, five equilibrium or be not divided into N on the vc3 direction ₁, N ₂, N ₃Part, form N _T=N ₁N ₂N ₃Individual lattice VCSPTN (i, j, k), and i=0 wherein, 1,2 ... N ₁-1, j=0,1,2 ... N ₂-1, k=0,1,2 ... N ₃-1.

The concrete lattice method branch as shown in Figure 9 principle schematic of formatting is as follows:

MBC is divided into a plurality of cuboid lattices, the even or non-homogeneous N that is divided on the VC1 direction ₁Section; Even or the non-homogeneous N that is divided on the VC2 direction ₂Section; Even or the non-homogeneous N that is divided on the VC3 direction ₃Section.Therefore MBC is divided into N _T=N ₁N ₂N ₃Individual lattice.Concrete division rule is as follows:

On the VC1 direction, interval [minvc1, maxvc1] evenly or anisotropically is divided into N ₁Individual subinterval, branch VC1 coordinate is dp ¹(i) (i=1,2 ... .N ₁-1), dp ¹(0)=and minvc1, dp ¹(N ₁)=maxvc1.So i (i=0,1,2 ... .N ₁-1) individual subinterval is [dp ¹(i), dp ¹(i+1)].

On the VC2 direction, interval [minvc2, maxvc2] evenly or anisotropically is divided into N ₂Individual subinterval, branch VC2 coordinate is dp ²(j) (j=1,2 ... .N ₂-1), dp ²(0)=and minvc2, dp ²(N ₂)=maxvc2. so j (j=0,1,2 ... .N ₂-1) individual subinterval is [dp ²(j), dp ²(j+1)].

On the VC3 direction, interval [minvc3, maxvc3] evenly or anisotropically is divided into N ₃Individual subinterval, branch VC3 coordinate is dp ³(k) (k=1,2 ... .N ₃-1), dp ³(0)=and minvc3, dp ³(N ₃)=maxvc3. so k (k=0,1,2 ... .N ₃-1) individual subinterval is [dp ³(k), dp ³(k+1)].

Therefore, (k) (k) (VCSPTN is the abbreviation of VCS ParTitioN) is subinterval [dp to individual lattice VCSPTN for i, j for i, j ¹(i), dp ¹(i+1)], [dp ²(j), dp ²(j+1)], [dp ³(k), dp ³(k+1)] cartesian product (Cartesian).Therefore there is following relational expression:

VCSPTN(i，j，k)＝

{(vc1，vc2，vc3)|dp ¹(i)≤vc1≤dp ¹(i+1)，dp ²(j)≤vc2≤dp ²(j+1)，dp ³(k)≤vc3≤dp3(k+1)}

..................................................................................[16]

Step S204 is cut apart the relation of formed each VCSPTN lattice and VAPS according to MBC, and all VCSPTN lattices are classified, and will merge with the lattice that VAPS exist non-NULL to occur simultaneously, and calculates its union UOP (Union of Partition).

MBC is cut apart formed each VCSPTN lattice, with the relation of VAPS three kinds of forms is arranged, and according to the relation of described MBC lattice and VAPS, all VCSPTN lattices of MBC is divided into three classes, and is specific as follows:

Certain lattice VCSPTN of MBC (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) be included among the VAPS, promptly satisfy: $\mathrm{VCSPTN}(i,j,k)\&SubsetEqual;\mathrm{VAPS}.$ These MBC lattices are belonged to first kind VCSPTN lattice.

Certain lattice VCSPTN of MBC (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) not exclusively be included among the VAPS, but with VAPS common factor (occur simultaneously and be not empty set) is arranged.Promptly $\mathrm{VCSPTN}(i,j,k)\&NotSubset;\mathrm{VAPS},$ VCSPTN(i，j，k)∩VAPS≠Φ。These MBC lattices are belonged to the second class VCSPTN lattice.

Certain lattice VCSPTN of MBC (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) with the common factor of VAPS be empty set.Be VCSPTN (i, j, k) ∩ VAPS=Φ.These MBC lattices are belonged to the 3rd class VCSPTN lattice.

It doesn't matter for obvious the 3rd class VCSPTN lattice and VAPS, so the present invention need not consider.The first kind and the second class VCSPTN lattice, its union (Union) is tied to form upright with UOP (Union of Partition) expression, obvious UOP as a spatial aggregation among the VCS just like ShiShimonoseki:

$\mathrm{VAPS}\&SubsetEqual;\mathrm{UOP}\&SubsetEqual;\mathrm{MBC}...............\left[17\right]$

Step S205, for each the lattice VCSPTN in the UOP set, the equation of determining its six faces of corresponding peimage lattice RGBPTN (abbreviation of RGB ParTitioN) in the RGB color space according to the math equation and the color direct transform function of its six faces.

For each VCSPTN in the UOP set (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) lattice, be a cuboid, therefore six faces are arranged, its label as shown in figure 10, six face equations are:

Face 1:

$\left\{\begin{array}{c}\mathrm{vc}3={\mathrm{dp}}^3\left(k\right)\\ {\mathrm{dp}}^1\left(i\right)\&le;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1{\&le;\mathrm{<figure-callout\; id="1"\; label="dp"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">dp</figure-callout>}}^1\&le;(i+1)\\ {\mathrm{dp}}^2\left(j\right)\&le;\mathrm{vc}2{\&le;\mathrm{dp}}^2(j+1)\end{array}\right................\left[18\right]$

Face 2:

$\left\{\begin{array}{c}\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1={\mathrm{dp}}^1(i+1)\\ {\mathrm{dp}}^2\left(j\right)\&le;\mathrm{vc}2\&le;{\mathrm{dp}}^2(j+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;\mathrm{vc}3{\&le;\mathrm{dp}}^3(k+1)\end{array}\right................\left[19\right]$

Face 3:

$\left\{\begin{array}{c}\mathrm{vc}2={\mathrm{dp}}^2(j+1)\\ {\mathrm{dp}}^1\left(i\right)\&le;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1{\&le;\mathrm{<figure-callout\; id="1"\; label="dp"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">dp</figure-callout>}}^1\&le;(i+1)\\ {\mathrm{dp}}^2\left(k\right)\&le;\mathrm{vc}3{\&le;\mathrm{dp}}^3(k+1)\end{array}\right................\left[20\right]$

Face 4:

$\left\{\begin{array}{c}\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1={\mathrm{dp}}^1\left(i\right)\\ {\mathrm{dp}}^2\left(j\right)\&le;\mathrm{vc}2{\&le;\mathrm{dp}}^2\&le;(j+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;\mathrm{vc}3{\&le;\mathrm{dp}}^3(k+1)\end{array}\right................\left[21\right]$

Face 5:

$\left\{\begin{array}{c}\mathrm{vc}2={\mathrm{dp}}^2\left(j\right)\\ {\mathrm{dp}}^1\left(i\right)\&le;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1{\&le;\mathrm{<figure-callout\; id="1"\; label="dp"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">dp</figure-callout>}}^1\&le;(i+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;\mathrm{vc}3{\&le;\mathrm{dp}}^3(k+1)\end{array}\right................\left[22\right]$

Face 6:

$\left\{\begin{array}{c}\mathrm{vc}3={\mathrm{dp}}^3(k+1)\\ {\mathrm{dp}}^1\left(i\right)\&le;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1{\&le;\mathrm{<figure-callout\; id="1"\; label="dp"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">dp</figure-callout>}}^1\&le;(i+1)\\ {\mathrm{dp}}^2\left(j\right)\&le;\mathrm{vc}2{\&le;\mathrm{dp}}^2(j+1)\end{array}\right................\left[23\right]$

For the VCSPTN lattice in the UOP set, all with the RGB color space in the RGBPTN lattice corresponding one by one.Relation as shown in figure 11.As seen from Figure 11, by color direct transform function T (.), (k) lattice is mapped to VCSPTN (i, j, k) lattice to RGBPTN for i, j; And by color counter-transformation function S (.), (k) lattice can be mapped to RGBPTN (i, j, k) lattice to VCSPTN for i, j.

Be defined in the rgb space and the lattice of first kind VCSPTN lattice correspondence is called first kind RGBPTN lattice; Be called the second class RGBPTN lattice with the lattice of the second class VCSPTN lattice correspondence.

If RGBPTN (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) be first kind lattice, (i, j k) are completely contained among the RGBUC, promptly RGBPTN so $\mathrm{RGBPTN}(i,j,k)\&SubsetEqual;\mathrm{RGBUC}.$

If RGBPTN (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) be the second class lattice, so RGBPTN (i, j k) not exclusively are included among the RGBUC, but and RGBUC common factor (occur simultaneously be not empty set) is arranged, promptly $\mathrm{RGBPTN}(i,j,k)\&NotSubset;\mathrm{RGBUC};$ RGBPTN(i，j，k)∩RGBUC≠Φ。

By above-mentioned analysis, can determine the approximate range of RGBPTN lattice.

For each RGBPTN lattice, six faces are also arranged, its label as shown in figure 12, each face of described RGBPTN lattice is corresponding with each face of VCSPTN lattice.For the RGBPTN lattice, if determined the equation of its six faces, then just determine the mathematical notation of this lattice fully, also just determine the spatial dimension of this lattice fully as a set.

RGBPTN (i, j, k) (i=0,1,2 ... N1, j=0,1,2 ... N2, k=0,1,2 ... N3) six of lattice face equations are as follows:

Face 1:

$\left\{\begin{array}{c}{t}_3(r,g,b)={\mathrm{dp}}^3\left(k\right)\\ {\mathrm{dp}}^1\left(i\right)\&le;t_1(r,g,b)\&le;{\mathrm{dp}}^1(i+1)\\ {\mathrm{dp}}^2\left(j\right)\&le;t_2(r,g,b)\&le;{\mathrm{dp}}^2(j+1)\end{array}\right................\left[24\right]$

Face 2:

$\left\{\begin{array}{c}{t}_1(r,g,b)={\mathrm{dp}}^1(i+1)\\ {\mathrm{dp}}^2\left(j\right)\&le;t_2(r,g,b)\&le;{\mathrm{dp}}^2(j+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;t_3(r,g,b)\&le;{\mathrm{dp}}^3(k+1)\end{array}\right...............\left[24\right]$

Face 3:

$\left\{\begin{array}{c}{t}_2(r,g,b)={\mathrm{dp}}^2(j+1)\\ {\mathrm{dp}}^1\left(i\right)\&le;t_1(r,g,b)\&le;{\mathrm{dp}}^1(i+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;t_3(r,g,b)\&le;{\mathrm{dp}}^3(k+1)\end{array}\right................\left[25\right]$

Face 4:

$\left\{\begin{array}{c}{t}_1(r,g,b)={\mathrm{dp}}^1\left(i\right)\\ {\mathrm{dp}}^2\left(j\right)\&le;t_2(r,g,b)\&le;{\mathrm{dp}}^2(j+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;t_3(r,g,b)\&le;{\mathrm{dp}}^3(k+1)\end{array}\right................\left[26\right]$

Face 5:

$\left\{\begin{array}{c}{t}_2(r,g,b)={\mathrm{dp}}^2\left(j\right)\\ {\mathrm{dp}}^1\left(i\right)\&le;t_1(r,g,b)\&le;{\mathrm{dp}}^1(i+1)\\ {\mathrm{dp}}^3\left(k\right)\&le;t_3(r,g,b)\&le;{\mathrm{dp}}^3(k+1)\end{array}\right................\left[27\right]$

Face 6:

$\left\{\begin{array}{c}{t}_3(r,g,b)={\mathrm{dp}}^3(k+1)\\ {\mathrm{dp}}^1\left(i\right)\&le;t_1(r,g,b)\&le;{\mathrm{dp}}^1(i+1)\\ {\mathrm{dp}}^2\left(j\right)\&le;t_2(r,g,b)\&le;{\mathrm{dp}}^2(j+1)\end{array}\right................\left[28\right]$

Step S206, according to the union UOP of all VCSPTN lattices that non-NULL occurs simultaneously being arranged with described VAPS, and the color counter-transformation function, obtain UOP and be integrated into peimage IIUOP in the RGB color space, and with it as the union of all RGBPTN lattices that non-NULL occurs simultaneously being arranged with described RGBUC.

The first kind and the second class VCSPTN lattice, the union of corresponding RGBPTN lattice in the RGB color space, being exactly UOP is mapped to set in the RGB color space by color counter-transformation function S (.), and this set is called IIUOP (Inverse Image of UOP, the inverse image of UOP).Obviously IIUOP comprises RGBUC fully, as shown in figure 13, satisfies the relation shown in the formula [29]:

$\mathrm{RGBUC}\&SubsetEqual;\mathrm{IIUOP}...............\left[29\right]$

Step S207 in described IIUOP set, determines the common factor of each RGBPTN lattice and described RGBUC, and the linear parameter of constructing the original gamma characteristic function on the described common factor represents, obtains corresponding linear gamma characteristic function; And determine described linear gamma characteristic function parameters according to the principle of setting, the principle of described setting is: by the caused distortion RGB of original gamma characteristic function color vector, and reach the target setting value by generalized mean distance between the vector between the caused distortion RGB of the described linear gamma characteristic function color vector.

Among the step S207,, on the common factor of each RGBPTN lattice and RGBUC, set up a mathematical optimization problem that contains the linear Gamma characterisitic function of this lattice parameter mainly according to minimum mean square error criterion; Find the solution this mathematical optimization problem then, determine described linear Gamma characterisitic function parameter.

Because Gamma characterisitic function G (.) is defined on the whole RGBUC, it is a nonlinear function, but on each RGBPTN lattice, can linearisation be similar to, promptly on optimum meaning, coming the approximate original Gamma characterisitic function that replaces with a linear function on the RGBPTN lattice.RGBPTN (i, j, k) (i=0,1,2 ... N ₁, j=0,1,2 ... N ₂, k=0,1,2 ... N ₃) on, this approximate representation is as follows:

r _d＝g _r(r _r)≈k _r(i)r _r+b _r(i)......................................................[30]

g _d＝g _g(g _r)≈k _g(j)g _r+b _g(j)................................................[31]

b _d＝g _b(b _r)≈k _b(k)b _r+b _b(k)...............................................[32]

For convenience, the back directly uses equal sign "=" to replace approximate number " ≈ ".Perhaps write as matrix form, shown in formula [33]:

$c_{\mathrm{dRGB}}=K(i,j,k)c_{\mathrm{rRGB}}+B(i,j,k)$

$\left[\begin{array}{ccc}{k}_r\left(i\right)& 0& 0\\ 0& k_g\left(j\right)& 0\\ 0& 0& k_b\left(k\right)\end{array}\right]c_{\mathrm{rRGB}}+\left[\begin{array}{c}{b}_r\left(i\right)\\ b_g\left(j\right)\\ b_b\left(k\right)\end{array}\right]...............\left[33\right]$

Wherein (i, j k) are one 3 * 3 matrix to K, and (i, j k) are one 3 * 1 column vector to B.

Suppose and obtained the RGBPTN lattice at the Gamma of Gamma link characterisitic function.If do not know, can adopt the method for apparatus measures to obtain.

According to following optiaml ciriterion determine in the Gamma characterisitic function parameter K (i, j, k) and B (i, j, k).

$\{K(i,j,k),B(i,j,k),i=\mathrm{1,2,3},....,{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">N</figure-callout>}}_1,j=\mathrm{1,2,3},....,N_2,k=\mathrm{1,2,3},....,N_3\}$

...........................................................[34]

The meaning of above mathematic(al) representation is, under the optimum meaning of mean square error, described K (i, j, k) and B (k) parameter should be to make target function (perhaps being called cost function) for i, j Get minimum value.Wherein,=(K (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)).

In target function (perhaps being called cost function), select the integration set to be RGBPTN (i, j, k) reason of ∩ RGBUC is: for first kind RGBPTN lattice, be contained in fully among the RGBUC, this time RGBPTN (i, j, k) ∩ RGBUC=RGBPTN (i, j, k), the integration set just in time is RGBPTN (i, j, k); For the second class RGBPTN lattice, not exclusively be contained among the RGBUC, but have non-NULL to occur simultaneously, at this moment, the integration set should be RGBPTN (i, j, k) with the common factor of RGBUC, i.e. RGBPTN (i, j, k) ∩ RGBUC.In sum, no matter for the first kind or the second class RGBPTN lattice, the integration set all is RGBPTN (i, j, k) ∩ RGBUC.

In the solution of above mathematical optimization problem, as long as can try to achieve the analytical form of G (.) function in the Gamma characterisitic function, just can utilize present prior art obtain parameter K (i, j, k) and B (i, j, k).

Step S208, determine the VCSPTN lattice of described RGBPTN lattice correspondence and the common factor of described VAPS, and according to the linear gamma characteristic function on the common factor of each RGBPTN that is constructed and described RGBUC, and setting principle, determine the linear Gamma correction function on the described common factor; Described setting principle is: the RGB color vector that the virtual color space color vector on the VCSPTN lattice after described linearity correction function make to be proofreaied and correct and the common factor of described VAPS obtains after by the color inverse transformation, and the generalized mean between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform is apart from reaching the target setting value.

In step S208, mainly according to minimum mean square error criterion, in the mathematical optimization problem of setting up the linear Gamma characterisitic function parameter on a linear Gamma correction function parameter that contains this lattice and the corresponding RGBPTN lattice on each VCSPTN lattice, find the solution this mathematical optimization problem then, determine described linear Gamma correction function parameter, thereby determine the linear Gamma correction function parameter on all VCSPTN lattices, and then determine linear Gamma correction function on all VCSPTN lattices according to described linear Gamma correction function parameter.

During linear Gamma correction function in determining the VAPS set on each VCSPTN lattice, basic thought is: in each VCSPTN lattice, all use same linear function to represent the Gamma characterisitic function; In the different VCSPTN lattices, linear function is also different.Still (k) lattice is that example describes for i, j with VCSPTN below.

In each VCSPTN lattice, represent Gamma correction function in this VCSPTN lattice with the segmented line shape function.Represent Gamma correction function (GC represents Gammacorrection) with GC (.).Concrete form is shown in formula [35]:

$c_{\mathrm{cVCS}}=\mathrm{GC}\left(c_{\mathrm{VCS}}\right)=\left[\begin{array}{c}{\mathrm{gc}}_1\left(c_{\mathrm{uVCS}}\right)\\ {\mathrm{gc}}_2\left(c_{\mathrm{uVCS}}\right)\\ {\mathrm{gc}}_3\left(c_{\mathrm{uVCS}}\right)\end{array}\right]$

$=\left[\begin{array}{c}{p}_1(i,j,k)c_{\mathrm{uVCS}}+q_1(i,j,k)={\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^1(i,j,k){\mathrm{vc}1}_u{+\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^2(i,j,k){\mathrm{vc}2}_u+{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^3(i,j,k){\mathrm{vc}3}_u+q_1(i,j,k)\\ p_2(i,j,k)c_{\mathrm{uVCS}}+q_2(i,j,k)=p_2^1(i,j,k){\mathrm{vc}1}_u+p_2^2(i,j,k){\mathrm{vc}2}_u+p_2^3(i,j,k){\mathrm{vc}3}_u+q_2(i,j,k)\\ p_3(i,j,k)c_{\mathrm{uVCS}}+q_3(i,j,k)=p_3^1(i,j,k){\mathrm{vc}1}_u+p_3^2(i,j,k){\mathrm{vc}2}_u+p_3^3(i,j,k){\mathrm{vc}3}_u+q_3(i,j,k)\end{array}\right]...\left[35\right]$

C wherein _UVCS∈ VCSPTN (i, j, k), and i=1,2 ... .N ₁, j=1,2 ... .N ₂, k=1,2 ... .N ₃

Wherein three of function G C (.) components are shown in formula [36]:

$\mathrm{GC}(.)=\left[\begin{array}{c}{\mathrm{gc}}_1(.)\\ {\mathrm{gc}}_2(.)\\ {\mathrm{gc}}_3(.)\end{array}\right]...............\left[36\right]$

Gc wherein ₁(.), gc ₂(.), gc ₃(.) represents to produce correction back VCS vector c respectively _CVCSVc1, vc2, vc3 component.

Wherein P (3 * 3 matrixes k) are shown in formula [37] for i, j:

$P(i,j,k)=\left[\begin{array}{c}{p}_1(i,j,k)\\ p_2(i,j,k)\\ p_3(i,j,k)\end{array}\right]=\left[\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^1(i,j,k){\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^2(i,j,k){\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^3(i,j,k)\\ p_2^1(i,j,k)p_2^2(i,j,k)p_2^3(i,j,k)\\ p_3^1(i,j,k)p_3^2(i,j,k)p_3^3(i,j,k)\end{array}\right]...............\left[37\right]$

Q (3 * 1 column vectors k) are shown in formula [38] for i, j:

$Q(i,j,k)=\left[\begin{array}{c}{q}_1(i,j,k)\\ q_2(i,j,k)\\ q_3(i,j,k)\end{array}\right]..................\left[38\right]$

Therefore, can obtain final Gamma correction function shown in formula [39]:

c _cVCS＝P(i，j，k)T(K(i，j，k)c _rRGB+B(i，j，k))+Q(i，j，k)........................[39]

Wherein, c _RRGB∈ RGBPTN (i, j, k) ∩ RGBUC, i=1,2 ... .N ₁, j=1,2 ... .N ₂, k=1,2 ... .N ₃

(k) (k) parameter, the method for use are by weighing the VCS color vector c after proofreading and correct to P in definite below Gamma correction function for i, j for parameter and Q for i, j _CVCSDo not have the size of the error of the VCS color vector under the situation of Gamma distortion with hypothesis, and, determine P (i, j, k) parameter and Q (i, j, k) parameter according to this error principle of hour definite Gamma correction function.

Supposing not have the VCS color vector under the situation of Gamma distortion is T (crRGB).Concrete errors of form, the size that can define the error of VCS color vector according to statistical signal process field mean-square error criteria commonly used is shown in formula [40]:

(c _cVCS-T(c _rRGB)) ^T(c _cVCS-T(c _rRGB)).......................................[40]

Therefore, on whole RGB color space, a kind of accumulation of this error is average in other words, can define with the integrated form shown in formula [41]:

$\mathrm{MSE}={\&Integral;}_{c_{\mathrm{cRGB}\&Element;{\left[\mathrm{0,1}\right]}^3}}{(c_{\mathrm{cVCS}}-T\left(c_{\mathrm{rRGB}}\right))}^T(c_{\mathrm{cVCS}}-T\left(c_{\mathrm{rRGB}}\right)){\mathrm{dc}}_{\mathrm{rRGB}}$

..........................................................[41]

Wherein MSE represents mean square error, and English is Mean Square Error.

Because each RGB lattice RGBPTN (i, j, k) with the VCS spatial dimension in the UOP set in each lattice VCSPTN (i, j, k) corresponding one by one, so, be defined in RGBPTN (i, j, k) the Gamma characterisitic function on is defined within VCSPTN, and (Gamma correction function k) is proofreaied and correct for i, j.The Gamma characterisitic function of other lattice and Gamma correction function do not influence RGBPTN (i, j, k) and VCSPTN (i, j, k).Therefore each VCSPTN (k) can independently determine for i, j by the Gamma correction function of lattice, only depend on RGBPTN (i, j, k) Gamma characterisitic function on the lattice, perhaps only depend on RGBPTN (i, j, the k) linear expression of Gamma characterisitic function on the lattice more strictly speaking.

For the second class lattice, consider that the second class lattice not exclusively is included in the VAPS set, and the Gamma correction function should be defined in the VAPS set.Therefore, for each lattice, at first to determine itself and VAPS intersection of sets collection.

(k) reason of ∩ VAPS is for i, j, for first kind VCSPTN lattice, is contained in the VAPS set fully for VCSPTN to select the integration set, have this time: VCSPTN (i, j, k) ∩ VAPS=VCSPTN (i, j, k), i.e. integration set just in time be VCSPTN (I, j, k); For the second class VCSPTN lattice, not exclusively be contained in the VAPS set, but have non-NULL to occur simultaneously, integration set in this time should be that (I, j is k) with common factor VCSPTN (i, j, k) the ∩ VAPS of VAPS for VCSPTN.In sum, no matter for the first kind or the second class VCSPTN, the integration set all is VCSPTN (i, j, k) ∩ VAPS.

According to this principle, come to determine the Gamma correction function by lattice ground.Because all lattices have constituted whole UOP set, as long as determined that so (k) the Gamma correction function on the lattice just can be determined the Gamma correction function in whole UOP set to each VCSPTN for i, j.

The criterion of determining the Gamma correction function institute foundation on each lattice is the minimum mean square error criterion of formula [34] definition, according to this principle, determine P (i, j, k) 3 * 3 matrixes of parameter, shown in formula [42]:

$P(i,j,k)=\left[\begin{array}{c}{p}_1(i,j,k)\\ p_2(i,j,k)\\ p_3(i,j,k)\end{array}\right]=\left[\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^1(i,j,k){\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^2(i,j,k){\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">p</figure-callout>}}_1^3(i,j,k)\\ p_2^1(i,j,k)p_2^2(i,j,k)p_2^3(i,j,k)\\ p_3^1(i,j,k)p_3^2(i,j,k)p_3^3(i,j,k)\end{array}\right].................\left[42\right]$

And, Q (i, j, k) 3 * 1 column vectors of parameter, shown in formula [43]:

$Q(i,j,k)=\left[\begin{array}{c}{q}_1(i,j,k)\\ q_2(i,j,k)\\ q_3(i,j,k)\end{array}\right]...........................\left[43\right]$

Make $\mathrm{MSE}={\&Integral;}_{c_{\mathrm{cRGB}\&Element;{\left[\mathrm{0,1}\right]}^3}}{(c_{\mathrm{cVCS}}-T\left(c_{\mathrm{rRGB}}\right))}^T(c_{\mathrm{cVCS}}-T\left(c_{\mathrm{rRGB}}\right)){\mathrm{dc}}_{\mathrm{rRGB}}$ Minimum.

Formula [39] substitution formula [41] is obtained:

$\mathrm{MSE}={\&Integral;}_{c_{\mathrm{cRGB}\&Element;{\left[\mathrm{0,1}\right]}^3}}\begin{array}{c}(P(i,j,k)T(K(i,j,k)c_{\mathrm{rRGB}}+B(i,j,k))+Q(i,j,k)-T\left(c_{\mathrm{rRGB}}\right){)}^T\\ (P(i,j,k)T(K(i,j,k)c_{\mathrm{rRGB}}+B(i,j,k))+Q(i,j,k)-T\left(c_{\mathrm{rRGB}}\right))\end{array}{\mathrm{dc}}_{\mathrm{rRGB}}$

........................................................[44]

So P _Opt(i, j, k), Q _Opt(k) definite forwarding of parameter is the mathematical optimization problem of classics, that is: for i, j

$\{P_{\mathrm{Opt}}(i,j,k),Q_{\mathrm{Opt}}(i,j,k)\}=$

$\underset{P(i,j,k),Q(i,j,k)}{\arg \min }{\&Integral;}_{c_{\mathrm{cRGB}\&Element;{\left[\mathrm{0,1}\right]}^3}}\begin{array}{c}(P(i,j,k)T(K(i,j,k)c_{\mathrm{rRGB}}+B(i,j,k))+Q(i,j,k)-T\left(c_{\mathrm{rRGB}}\right){)}^T\\ (P(i,j,k)T(K(i,j,k)c_{\mathrm{rRGB}}+B(i,j,k))+Q(i,j,k)-T\left(c_{\mathrm{rRGB}}\right))\end{array}{\mathrm{dc}}_{\mathrm{rRGB}}$

..................................................................[45]

The method of employing mathematical optimization is tried to achieve the globally optimal solution P in the formula [45] _Opt(i, j, k), Q _Opt(k), (k) the Gamma correction function on the lattice just can be determined the Gamma correction function in whole UOP set for i, j so just can to determine each VCSPTN for i, j.Wherein Opt is the abbreviation of English Optimal (optimum).The method for solving of this part belongs to prior art, does not describe in this patent scope.

Because in the video communication process, the video information that sends to receiving terminal (a plurality of receiving terminals can be arranged) from transmitting terminal is the sampled value sequence that some color signal sampled values form in fact.For video, comprise the image of a sequence, and every two field picture comprises a plurality of pixels, each pixel is corresponding to a color vector signals sampling value.Therefore after all, video can be expressed as the color vector signals sampling value of a sequence.In the video communication process, obtain each the original RGB vector signal of telecommunication sampled value that comprises in the video information, and after being converted into the virtual color space vector signal of telecommunication in the virtual color space, obtain the virtual color space vector of corresponding VCS signal of telecommunication sampled value, utilize the Gamma correction function on each VCSPTN lattice that said process determines that described virtual color space vector signal of telecommunication sampled value is carried out Gamma correction then, specific as follows:

Step S209, in the video communication process, obtain each the original RGB vector signal of telecommunication sampled value that comprises in the video information, and after being converted into virtual color space vector signal in the virtual color space, obtain the corresponding virtual color space vector signal of VCS sampled value, judge the VCSPTN lattice that the virtual color space vector signal of described VCS sampled value is belonged to then.

In step S209, with described virtual color space vector signal sampled value corresponding component on three different reference axis, compare with the coordinate range of each VCSPTN lattice, if satisfy:

$\left\{\begin{array}{c}{\mathrm{dp}}^1\left(i_0\right)\&le;\mathrm{<figure-callout\; id="1"\; label="vc"\; filenames="A20071000098900412.png"\; state="\{\{state\}\}">vc</figure-callout>}1\&le;{\mathrm{dp}}^1(i_0+1)\\ {\mathrm{dp}}^2\left(j_0\right)\&le;\mathrm{vc}2\&le;{\mathrm{dp}}^2(j_0+1)\\ {\mathrm{dp}}^3\left(k_0\right)\&le;\mathrm{vc}3\&le;{\mathrm{dp}}^3(k_0+1)\end{array}\right...........................\left[46\right]$

Determine that so described VCS color vector belongs to VCSPTN (i ₀, j ₀, k ₀) lattice.That is to say, when described virtual color space vector signal sampled value at component corresponding on three different reference axis all in the coordinate range of certain VCSPTN lattice on the respective coordinates axle time, determine that then described virtual color space vector signal sampled value belongs to described VCSPTN lattice.

Step S210 is according to determined VCSPTN (i ₀, j ₀, k ₀) linear Gamma correction function (shown in formula [47]) on the lattice, described VCS vector signal sampled value is proofreaied and correct.

c _cVCS＝P(i ₀，j ₀，k ₀)c _uVCS+Q(i ₀，j ₀，k ₀)........................................[47]

The next one in the video information is not proofreaied and correct VCS signals sampling value repeating step S209 and step S210, all dispose up to all sampled values.

Wherein, when utilizing linear Gamma correction function that signal is carried out Gamma correction, can adopt the bearing calibration of present standard, making final signal input/output relation is linear relationship.

Second embodiment provided by the invention is a kind of Gamma correction system that is used for video communication, and its structure comprises as shown in figure 14: lattice determining unit, linear Gamma correction function determining unit, lattice matching unit and gammate.Wherein, described lattice determining unit comprises that first spatial dimension determines that subelement, second spatial dimension determine that subelement and lattice determine subelement; Described second spatial dimension determines that subelement comprises that the VAPS coordinate determines that subelement, MBC coordinate determine that subelement and MBC spatial dimension determine subelement.Wherein, described linear Gamma correction function determining unit comprises that linear gamma characteristic function determines that subelement and linear Gamma correction function determine subelement; Wherein said linear gamma characteristic function determines that subelement comprises RGBPTN lattice set determination module and linear gamma characteristic function determination module; Wherein, described lattice matching unit comprises comparison subelement and the definite subelement of lattice ownership.

Signal transitive relation in the system between each unit is as follows:

Described lattice determining unit allows the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space in the virtual color space, VAPS is divided format, and forms a plurality of VCSPTN lattices, and concrete processing procedure is as follows:

Described first spatial dimension is determined subelement, according to the RGB color space being carried out the math equation that six faces of resulting RGB unit cube RGBUC are handled in normalization, and the color counter-transformation function, determines the equation of six faces of VAPS; Six faces of described VAPS are for surrounding six space curved surfaces of described VAPS; And, according to the equation of six faces of described VAPS, determine the spatial dimension of VAPS; Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Second spatial dimension determines that subelement determines the spatial dimension of the VAPS that subelement is determined according to described first spatial dimension, determines the spatial dimension that its minimum comprises cuboid MBC; During concrete the processing, determine subelement, determine its maximum coordinates and min coordinates on three different change in coordinate axis direction according to the spatial dimension of described VAPS by the VAPS coordinate; And, determine that by the MBC coordinate subelement determines that according to described VAPS coordinate minimum that VAPS that subelement determines determines described VAPS in maximum coordinates on described three different change in coordinate axis direction and min coordinates comprises maximum coordinates and the min coordinates on three different change in coordinate axis direction of cuboid MBC in correspondence; And, determine subelement by the MBC spatial dimension, maximum coordinates and the min coordinates spatial dimension of determining described MBC of the MBC that is determined according to described MBC coordinate determining unit on three different change in coordinate axis direction of correspondence.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Described lattice determines that subelement determines that with described second spatial dimension MBC that subelement is determined is divided into a plurality of VCSPTN lattices, obtains a plurality of VCSPTN lattices of described VAPS.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Linear Gamma correction function on each the VCSPTN lattice that obtains after described linear Gamma correction function determining unit structure is handled by described lattice determining unit, the RGB color vector that described linear Gamma correction function makes the virtual color space color vector on the VCSPTN lattice after the correction obtain by the color inverse transformation, and the generalized mean distance reaches the target setting value between the vector between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform.Its processing procedure determines that by described linear gamma characteristic function subelement and linear Gamma correction function determine that subelement realizes:

Described linear gamma characteristic function determines that subelement passes through the RGBPTN lattice set determination module in it, according to the math equation of six faces of described VCSPTN lattice and the spatial dimension that color direct transform function is determined its corresponding peimage lattice RGBPTN in the RGB color space; And, according to the union UOP of all VCSPTN lattices that non-NULL occurs simultaneously being arranged with described VAPS, and the color counter-transformation function, obtain UOP and be integrated into peimage IIUOP in the RGB color space, and with it as the union of all RGBPTN lattices that non-NULL occurs simultaneously being arranged with described RGBUC; And, in described IIUOP set, determine the common factor of each RGBPTN lattice and described RGBUC; Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Then, construct the linear parameter of the original gamma characteristic function on the common factor that described RGBPTN lattice set determination module determines by described linear gamma characteristic function determination module and represent, obtain corresponding linear gamma characteristic function; And determine described linear gamma characteristic function parameters according to the principle of setting, the principle of described setting is: by the caused distortion RGB of original gamma characteristic function color vector, and reach the target setting value by generalized mean distance between the vector between the caused distortion RGB of the described linear gamma characteristic function color vector; Describedly be meant mean square error based on the generalized mean between vector distance.Described target setting value is described based on the generalized mean minimum value and value between vector.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Described linear Gamma correction function determines that subelement determines the VCSPTN lattice of described RGBPTN lattice correspondence and the common factor of described VAPS, and determine linear gamma characteristic function on the common factor of each RGBPTN that subelement is constructed and described RGBUC according to described linear gamma characteristic function, and setting principle, determine the linear Gamma correction function on the described common factor; Described setting principle is: the RGB color vector that the virtual color space color vector on the VCSPTN lattice after described linearity correction function make to be proofreaied and correct and the common factor of described VAPS obtains after by the color inverse transformation, and the generalized mean between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform is apart from reaching the target setting value.Describedly be meant mean square error based on the generalized mean between vector distance.Described target setting value is described based on the generalized mean minimum value and value between vector.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Described lattice matching unit obtains the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that comprises in the video information, and by comparing the component of subelement with described virtual color space vector signal sampled value correspondence on three different reference axis, compare with the coordinate range of each VCSPTN lattice, and send comparative result to described lattice ownership definite subelement; Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Determine subelement when described lattice ownership and determine that comparative result is described virtual color space vector signal sampled value at component corresponding on three different reference axis all in the coordinate range of certain VCSPTN lattice on the respective coordinates axle time, determines that then described virtual color space vector signal sampled value belongs to described VCSPTN lattice.Associated description among the comparative approach of coordinate and first embodiment is identical, is not described in detail here.

Described gammate calls the linear Gamma correction function on the VCSPTN lattice that described lattice matching unit matches, and resulting each virtual color space vector signal sampled value is carried out Gamma correction.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

The 3rd embodiment provided by the invention is a kind of Gamma correction device, and its structure comprises lattice determining unit and linear Gamma correction function determining unit as shown in figure 15.Wherein said lattice determining unit comprises that first spatial dimension determines that subelement, second spatial dimension determine that subelement and lattice determine subelement; Described second spatial dimension determines that subelement comprises that the VAPS coordinate determines that subelement, MBC coordinate determine that subelement and MBC spatial dimension determine subelement.Described linear Gamma correction function determining unit comprises that linear gamma characteristic function determines that subelement and linear Gamma correction function determine subelement.Wherein said linear gamma characteristic function determines that subelement comprises RGBPTN lattice set determination module and linear gamma characteristic function determination module.

Described lattice determining unit allows the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space in the virtual color space, VAPS is divided format, and forms a plurality of VCSPTN lattices, and concrete processing procedure is as follows:

Described first spatial dimension is determined subelement, according to the RGB color space being carried out the math equation that six faces of resulting RGB unit cube RGBUC are handled in normalization, and the color counter-transformation function, determines the equation of six faces of VAPS; Six faces of described VAPS are for surrounding six space curved surfaces of described VAPS; And, according to the equation of six faces of described VAPS, determine the spatial dimension of VAPS; Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Second spatial dimension determines that subelement determines the spatial dimension of the VAPS that subelement is determined according to described first spatial dimension, determines the spatial dimension that its minimum comprises cuboid MBC; During concrete the processing, determine subelement, determine its maximum coordinates and min coordinates on three different change in coordinate axis direction according to the spatial dimension of described VAPS by the VAPS coordinate; And, determine that by the MBC coordinate subelement determines that according to described VAPS coordinate minimum that VAPS that subelement determines determines described VAPS in maximum coordinates on described three different change in coordinate axis direction and min coordinates comprises maximum coordinates and the min coordinates on three different change in coordinate axis direction of cuboid MBC in correspondence; And, determine subelement by the MBC spatial dimension, maximum coordinates and the min coordinates spatial dimension of determining described MBC of the MBC that is determined according to described MBC coordinate determining unit on three different change in coordinate axis direction of correspondence.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Described lattice determines that subelement determines that with described second spatial dimension MBC that subelement is determined is divided into a plurality of VCSPTN lattices, obtains a plurality of VCSPTN lattices of described VAPS.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Linear Gamma correction function on each the VCSPTN lattice that obtains after linear Gamma correction function determining unit structure on the described lattice is handled by described lattice determining unit, the RGB color vector that described linear Gamma correction function makes the virtual color space color vector on the VCSPTN lattice after the correction obtain by the color inverse transformation, and the generalized mean distance reaches the target setting value between the vector between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform.Its processing procedure determines that by described linear gamma characteristic function subelement and linear Gamma correction function determine that subelement realizes:

Described linear gamma characteristic function determines that subelement passes through the RGBPTN lattice set determination module in it, according to the math equation of six faces of described VCSPTN lattice and the spatial dimension that color direct transform function is determined its corresponding peimage lattice RGBPTN in the RGB color space; And, according to the union UOP of all VCSPTN lattices that non-NULL occurs simultaneously being arranged with described VAPS, and the color counter-transformation function, obtain UOP and be integrated into peimage IIUOP in the RGB color space, and with it as the union of all RGBPTN lattices that non-NULL occurs simultaneously being arranged with described RGBUC; And, in described IIUOP set, determine the common factor of each RGBPTN lattice and described RGBUC; Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Then, construct the linear parameter of the original gamma characteristic function on the common factor that described RGBPTN lattice set determination module determines by described linear gamma characteristic function determination module and represent, obtain corresponding linear gamma characteristic function; And determine described linear gamma characteristic function parameters according to the principle of setting, the principle of described setting is: by the caused distortion RGB of original gamma characteristic function color vector, and reach the target setting value by generalized mean distance between the vector between the caused distortion RGB of the described linear gamma characteristic function color vector; Describedly be meant mean square error based on the generalized mean between vector distance.Described target setting value is described based on the generalized mean minimum value and value between vector.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

Described linear Gamma correction function determines that subelement determines the VCSPTN lattice of described RGBPTN lattice correspondence and the common factor of described VAPS, and determine linear gamma characteristic function on the common factor of each RGBPTN that subelement is constructed and described RGBUC according to described linear gamma characteristic function, and setting principle, determine the linear Gamma correction function on the described common factor; Described setting principle is: the RGB color vector that the virtual color space color vector on the VCSPTN lattice after described linearity correction function make to be proofreaied and correct and the common factor of described VAPS obtains after by the color inverse transformation, and the generalized mean between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform is apart from reaching the target setting value.Describedly be meant mean square error based on the generalized mean between vector distance.Described target setting value is described based on the generalized mean minimum value and value between vector.Associated description among the specific implementation process and first embodiment is identical, is not described in detail here.

The specific embodiments that is provided by the embodiment of the invention described above as can be seen, the present invention at first allows the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space in the virtual color space, VAPS divided format, form a plurality of VCSPTN lattices; Construct the linear Gamma correction function on each VCSPTN lattice then; Then obtain the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that comprises in the video information, and determine its VCSPTN lattice that is belonged to; And, resulting each virtual color space vector signal sampled value is carried out Gamma correction according to the linear Gamma correction function on the described VCSPTN lattice.Therefore by the present invention, can in the shades of colour space that the intermediate treatment link of video communication process is used, implement Gamma correction, thereby the range of application of Gamma correction is enlarged, solved the current difficult problem that on video communication intermediate treatment link, can't implement Gamma correction.

In addition, the scheme that the embodiment of the invention described above provides can be widely used in the shades of colour space of using in the intermediate treatment link in the video communication process, thereby provides unified method for implement Gamma correction in the different colours space.

Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, if of the present invention these are revised and modification belongs within the scope of claim of the present invention and equivalent technologies thereof, then the present invention also is intended to comprise these changes and modification interior.

Claims (21)

1. a gamma revision method is characterized in that, comprising:

Allow the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space in the virtual color space, VAPS is divided format, form a plurality of VCSPTN lattices;

Construct the linear Gamma correction function on each VCSPTN lattice;

Obtain the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that comprises in the video information, and determine the VCSPTN lattice that described virtual color space vector signal sampled value is belonged to;

According to the linear Gamma correction function on the described VCSPTN lattice, resulting each virtual color space vector signal sampled value is carried out Gamma correction.

2. the method for claim 1 is characterized in that, described minimum according to the VAPS in the virtual color space comprises cuboid MBC, VAPS is divided format, and forms the process of a plurality of VCSPTN lattices, specifically comprises:

According to the RGB color space being carried out the math equation that six faces of resulting RGB unit cube RGBUC are handled in normalization, and the color counter-transformation function, determine the equation of six faces of VAPS; Six faces of described VAPS are for surrounding six space curved surfaces of described VAPS;

According to the equation of six faces of described VAPS, determine the spatial dimension of VAPS, determine the spatial dimension that its minimum comprises cuboid MBC according to the spatial dimension of described VAPS then;

MBC is divided into a plurality of VCSPTN lattices, obtains a plurality of VCSPTN lattices of described VAPS.

3. method as claimed in claim 2 is characterized in that, described spatial dimension according to described VAPS is determined the process that its minimum comprises the spatial dimension of cuboid MBC, specifically comprises:

Determine its maximum coordinates and min coordinates on three different change in coordinate axis direction according to the spatial dimension of described VAPS, the minimum of determining described VAPS in maximum coordinates on described three different change in coordinate axis direction and min coordinates according to described VAPS comprises maximum coordinates and the min coordinates on three different change in coordinate axis direction of cuboid MBC in correspondence then;

According to maximum coordinates and the min coordinates spatial dimension of determining described MBC of the MBC that is determined on three different change in coordinate axis direction of correspondence.

4. method as claimed in claim 3 is characterized in that, and is described according to the maximum coordinates of the MBC that is determined on three different change in coordinate axis direction of correspondence and the process of the min coordinates spatial dimension of determining described MBC, specifically comprises:

When described virtual color space adopts left-handed coordinate system, according to the coordinate on the MBC that is determined summit after the maximum coordinates on three different change in coordinate axis direction of correspondence and min coordinates obtain described MBC lower-left and the coordinate on upper right preceding summit; According to the coordinate on summit behind the described lower-left and the coordinate on upper right preceding summit, determine the coordinate on all the other summits of described MBC; Or,

When described virtual color space adopts right-handed coordinate system, according to the coordinate on the MBC that is determined summit after the maximum coordinates on three different change in coordinate axis direction of correspondence and min coordinates obtain described MBC bottom right and the coordinate on upper left preceding summit; According to the coordinate on summit behind the described bottom right and the coordinate on upper left preceding summit, determine the coordinate on all the other summits of described MBC.

5. the method for claim 1 is characterized in that, the process of the linear Gamma correction function on each VCSPTN lattice of described structure specifically comprises:

According to described VCSPTN lattice and described VAPS, and the color counter-transformation function, determine the common factor of each RGBPTN lattice and described RGBUC, and construct the linear gamma characteristic function on the common factor of each RGBPTN and described RGBUC;

Determine the VCSPTN lattice of described RGBPTN lattice correspondence and the common factor of described VAPS, and according to the linear gamma characteristic function on the common factor of each RGBPTN that is constructed and described RGBUC, and, the RGB color vector that obtains by the color inverse transformation based on the virtual color space color vector on the VCSPTN lattice that makes by linear Gamma correction function after the correction, and generalized mean distance between the vector between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform is constructed the linear Gamma correction function on the described common factor.

6. method as claimed in claim 5, it is characterized in that, described according to described VCSPTN lattice and described VAPS, and color counter-transformation function, determine the common factor of each RGBPTN lattice and described RGBUC, and construct the process of the linear gamma characteristic function on the common factor of each RGBPTN and described RGBUC, specifically comprise:

The spatial dimension of determining its corresponding peimage lattice RGBPTN in the RGB color space according to the math equation and the color direct transform function of six faces of described VCSPTN lattice;

According to the union UOP of all VCSPTN lattices that non-NULL occurs simultaneously being arranged with described VAPS, and color counter-transformation function, obtain UOP and be integrated into peimage IIUOP in the RGB color space, and with it as the union of all RGBPTN lattices that non-NULL occurs simultaneously being arranged with described RGBUC;

In described IIUOP set, determine the common factor of each RGBPTN lattice and described RGBUC, and the linear parameter of constructing the original gamma characteristic function on the described common factor represents, obtain corresponding linear gamma characteristic function; And according to the caused distortion RGB of original gamma characteristic function color vector, and by the definite described linear gamma characteristic function parameters of generalized mean distance between the vector between the caused distortion RGB of the described linear gamma characteristic function color vector.

7. as claim 1,5 or 6 described methods, it is characterized in that, describedly be meant mean square error based on the generalized mean between vector distance.

8. the method for claim 1 is characterized in that, the described process of determining the VCSPTN lattice that described virtual color space vector signal sampled value is belonged to specifically comprises:

Component with described virtual color space vector signal sampled value correspondence on three different reference axis, compare with the coordinate range of each VCSPTN lattice, and when described virtual color space vector signal sampled value at component corresponding on three different reference axis all in the coordinate range of certain VCSPTN lattice on the respective coordinates axle time, determine that then described virtual color space vector signal sampled value belongs to described VCSPTN lattice.

9. a Gamma correction system that is used for video communication is characterized in that, comprising:

The lattice determining unit is used for allowing the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space of virtual color space, VAPS is divided format, and forms a plurality of VCSPTN lattices;

Linear Gamma correction function determining unit is used to construct the linear Gamma correction function on each VCSPTN lattice;

The lattice matching unit is used for obtaining the pairing virtual color space vector signal sampled value of each original RGB vector signal of telecommunication sampled value that video information comprises, and determines its VCSPTN lattice that is belonged to;

Gammate is used for according to the linear Gamma correction function on the described VCSPTN lattice, and resulting each virtual color space vector signal sampled value is carried out Gamma correction.

10. system as claimed in claim 9 is characterized in that, described lattice determining unit comprises:

First spatial dimension is determined subelement, is used for according to the RGB color space is carried out the math equation that six faces of resulting RGB unit cube RGBUC are handled in normalization, and the color counter-transformation function, determines the equation of six faces of VAPS; Six faces of described VAPS are for surrounding six space curved surfaces of described VAPS; And, according to the equation of six faces of described VAPS, determine the spatial dimension of VAPS;

Second spatial dimension is determined subelement, is used for determining according to described first spatial dimension spatial dimension of the VAPS that subelement is determined, determines the spatial dimension that its minimum comprises cuboid MBC; And,

Lattice is determined subelement, is used for described second spatial dimension is determined that the MBC that subelement is determined is divided into a plurality of VCSPTN lattices, obtains a plurality of VCSPTN lattices of described VAPS.

11. system as claimed in claim 10 is characterized in that, described second spatial dimension determines that subelement also comprises:

The VAPS coordinate is determined subelement, is used for determining its maximum coordinates and min coordinates on three different change in coordinate axis direction according to the spatial dimension of described VAPS; And,

MBC coordinate determining unit is used for determining that according to described VAPS coordinate minimum that VAPS that subelement determines determines described VAPS in maximum coordinates on described three different change in coordinate axis direction and min coordinates comprises maximum coordinates and the min coordinates on three different change in coordinate axis direction of cuboid MBC in correspondence; And,

The MBC spatial dimension is determined subelement, is used for the MBC that determined according to described MBC coordinate determining unit maximum coordinates and the min coordinates spatial dimension of determining described MBC on three different change in coordinate axis direction of correspondence.

12. system as claimed in claim 9 is characterized in that, described linear Gamma correction function determining unit comprises:

Linear gamma characteristic function determines that subelement and linear Gamma correction function determine subelement;

Described linear gamma characteristic function is determined subelement, be used for according to described VCSPTN lattice and described VAPS, and the color counter-transformation function, determine the common factor of each RGBPTN lattice and described RGBUC, and construct the linear gamma characteristic function on the common factor of each RGBPTN and described RGBUC;

Described linear Gamma correction function is determined subelement, be used for determining the VCSPTN lattice of described RGBPTN lattice correspondence and the common factor of described VAPS, and according to the linear gamma characteristic function on the common factor of each RGBPTN that is constructed and described RGBUC, and the RGB color vector that obtains by the color inverse transformation based on the virtual color space color vector on the VCSPTN lattice that makes by linear Gamma correction function after the correction, and generalized mean distance between the vector between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform is constructed the linear Gamma correction function on the described common factor.

13. system as claimed in claim 12 is characterized in that, described linear gamma characteristic function determines that subelement comprises:

RGBPTN lattice set determination module and linear gamma characteristic function determination module;

Described RGBPTN lattice set determination module is used for determining its spatial dimension at the corresponding peimage lattice RGBPTN of RGB color space according to the math equation and the color direct transform function of six faces of described VCSPTN lattice; And, according to the union UOP of all VCSPTN lattices that non-NULL occurs simultaneously being arranged with described VAPS, and the color counter-transformation function, obtain UOP and be integrated into peimage IIUOP in the RGB color space, and with it as the union of all RGBPTN lattices that non-NULL occurs simultaneously being arranged with described RGBUC; And, in described IIUOP set, determine the common factor of each RGBPTN lattice and described RGBUC;

Linear gamma characteristic function determination module, the linear parameter that is used to construct the original gamma characteristic function on the common factor that described RGBPTN lattice set determination module determines is represented, obtains corresponding linear gamma characteristic function; And according to by the caused distortion RGB of original gamma characteristic function color vector, and determine described linear gamma characteristic function parameters by generalized mean distance between the vector between the caused distortion RGB of the described linear gamma characteristic function color vector.

14. as claim 9,12 or 13 described systems, it is characterized in that, describedly be meant mean square error based on the generalized mean between vector distance.

15. system as claimed in claim 9 is characterized in that, described lattice matching unit also comprises:

Relatively subelement and lattice ownership is determined subelement;

Subelement relatively is used for described virtual color space vector signal sampled value corresponding component on three different reference axis, compares with the coordinate range of each VCSPTN lattice; And comparative result is fed back to described lattice ownership determine subelement;

Described lattice ownership is determined subelement, be used for when comparative result be described virtual color space vector signal sampled value at component corresponding on three different reference axis all in the coordinate range of certain VCSPTN lattice on the respective coordinates axle time, determine that described virtual color space vector signal sampled value belongs to described VCSPTN lattice.

16. a Gamma correction device is characterized in that, comprising:

The lattice determining unit is used for allowing the minimum of some set VAPS to comprise cuboid MBC according to the virtual color space of virtual color space, VAPS is divided format, and forms a plurality of VCSPTN lattices;

Linear Gamma correction function determining unit is used to construct the linear Gamma correction function on each VCSPTN lattice.

17. device as claimed in claim 16 is characterized in that, described lattice determining unit comprises:

First spatial dimension is determined subelement, is used for according to the RGB color space is carried out the math equation that six faces of resulting RGB unit cube RGBUC are handled in normalization, and the color counter-transformation function, determines the equation of six faces of VAPS; Six faces of described VAPS are for surrounding six space curved surfaces of described VAPS; And, according to the equation of six faces of described VAPS, determine the spatial dimension of VAPS;

Second spatial dimension is determined subelement; Determine the spatial dimension of the VAPS that subelement is determined then according to described first spatial dimension, determine the spatial dimension that its minimum comprises cuboid MBC; And,

Lattice is determined subelement, is used for described second spatial dimension is determined that the MBC that subelement is determined is divided into a plurality of VCSPTN lattices, obtains a plurality of VCSPTN lattices of described VAPS.

18. device as claimed in claim 17 is characterized in that, described second spatial dimension determines that subelement also comprises:

The VAPS coordinate is determined subelement, is used for determining its maximum coordinates and min coordinates on three different change in coordinate axis direction according to the spatial dimension of described VAPS; And,

The MBC coordinate is determined subelement, is used for determining that according to described VAPS coordinate minimum that VAPS that subelement determines determines described VAPS in maximum coordinates on described three different change in coordinate axis direction and min coordinates comprises maximum coordinates and the min coordinates on three different change in coordinate axis direction of cuboid MBC in correspondence; And,

The MBC spatial dimension is determined subelement, is used for the MBC that determined according to described MBC coordinate determining unit maximum coordinates and the min coordinates spatial dimension of determining described MBC on three different change in coordinate axis direction of correspondence.

19. device as claimed in claim 16 is characterized in that, described linear Gamma correction function determining unit comprises:

Linear gamma characteristic function determines that subelement and linear Gamma correction function determine subelement;

Described linear gamma characteristic function is determined subelement, be used for according to described VCSPTN lattice and described VAPS, and the color counter-transformation function, determine the common factor of each RGBPTN lattice and described RGBUC, and construct the linear gamma characteristic function on the common factor of each RGBPTN and described RGBUC;

Described linear Gamma correction function is determined subelement, be used for determining the VCSPTN lattice of described RGBPTN lattice correspondence and the common factor of described VAPS, and according to the linear gamma characteristic function on the common factor of each RGBPTN that is constructed and described RGBUC, and the RGB color vector that obtains by the color inverse transformation based on the virtual color space color vector on the VCSPTN lattice that makes by linear Gamma correction function after the correction, and generalized mean distance between the vector between the original RGB color vector of the virtual color space color vector before obtaining proofreading and correct by the color direct transform is determined the linear Gamma correction function on the described common factor.

20. device as claimed in claim 19 is characterized in that, described linear gamma characteristic function determines that subelement comprises:

RGBPTN lattice set determination module and linear gamma characteristic function determination module;

Described RGBPTN lattice set determination module is used for determining its spatial dimension at the corresponding peimage lattice RGBPTN of RGB color space according to the math equation and the color direct transform function of six faces of described VCSPTN lattice; And, according to the union UOP of all VCSPTN lattices that non-NULL occurs simultaneously being arranged with described VAPS, and the color counter-transformation function, obtain UOP and be integrated into peimage IIUOP in the RGB color space, and with it as the union of all RGBPTN lattices that non-NULL occurs simultaneously being arranged with described RGBUC; And, in described IIUOP set, determine the common factor of each RGBPTN lattice and described RGBUC;

Linear gamma characteristic function determination module, the linear parameter that is used to construct the original gamma characteristic function on the common factor that described RGBPTN lattice set determination module determines is represented, obtains corresponding linear gamma characteristic function; And according to by the caused distortion RGB of original gamma characteristic function color vector, and determine described linear gamma characteristic function parameters by generalized mean distance between the vector between the caused distortion RGB of the described linear gamma characteristic function color vector.

21. as claim 16,19 or 20 described devices, it is characterized in that, describedly be meant mean square error based on the generalized mean between vector distance.

Images