WO2007147363A1

WO2007147363A1 - A method and an apparatus for correcting the gamma characteristic of the video communication

Info

Publication number: WO2007147363A1
Application number: PCT/CN2007/070121
Authority: WO
Inventors: Zhong Luo
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2006-06-15
Filing date: 2007-06-15
Publication date: 2007-12-27
Also published as: CN101047869B; CN101047869A

Abstract

A method and an apparatus for correcting the gamma characteristic of the video communication are provided, the method comprises: obtaining the luminance histogram of the output signal, converting the luminance histogram of the output luminance signal into the luminance distribution probability density function of the output luminance signal, determining the extremal points of the luminance distribution probability density function of the output luminance signal; building the mathematic relation between the extremal points of the luminance distribution probability density function of the output luminance signal and the extremal points of the luminance distribution probability density function of the input luminance signal; converting the mathematic relation between the luminance distribution probability density functions of the input and the output luminance signal respectively and the Gamma characteristic function into the mathematic relation between the luminance distribution probability density function of the output luminance signal and the Gamma characteristic function on the extremal points by using the mathematic relation on the extremal points; solving the converted mathematic relation to determine the Gamma characteristic parameter and executing the Gamma correction in the Gamma step. The method for determining the entirely blind Gamma characteristic parameter of the embodiment makes the Gamma correction more convenient, and extends the applying field of the Gamma correction.

Description

Method and device for correcting gamma characteristic of video communication

Technical field

The present invention relates to the field of video communication technologies, and in particular, to a method and apparatus for correcting gamma characteristics of video communication. Background of the invention

For video communication, the distortion of the luminance signal caused by the Gamma nonlinear problem caused by each link

(Distortion) is an important factor affecting the end user experience.

At present, methods and techniques for improving the end-to-end user experience mainly focus on pre-processing (post-processing) related to ensuring network QoS and video compression coding. There is a lack of attention and systematic solutions to the problem of brightness distortion caused by Gamma characteristics.

The following is a brief introduction to the Gamma feature.

The process of video communication is: optical signals that need to be transmitted, such as people, backgrounds, files, etc., enter the video communication terminal (hereinafter referred to as the terminal) such as a camera/camera, etc., are converted into digital image signals by A/D, and then pressed. The code is reduced, transmitted, and reaches the other terminal. Then, it is decompressed and decoded to be restored to a digital image signal, and then displayed on the display device, and finally becomes an optical signal that is perceived by the human eye. In the above process, the image brightness signal has gone through multiple steps. By definition, the Gamma property is a link where the input-output relationship of the image luminance signal is not linear, but a nonlinear relationship. The image luminance signal (Luminance) here is a generalized luminance signal, that is, the initial optical signal, the electrical signal, and then the digitized image luminance/gray signal, and the signal of each phase contains the information of the image luminance signal. Therefore, in a broad sense, the image brightness signal passes through multiple links.

A general model of the Gamma characteristics of a single link is shown in Figure 1.

In Figure 1, the nonlinear relationship between the input luminance signal and the output luminance signal can be expressed as: L. _Ut = G (L _in ) , where L _out is the output luminance signal, L _in is the input luminance signal, and the function G ( . ) is a nonlinear function.

An example of a typical Gamma characteristic is shown in Figure 2.

The numbers indicated in each of the blocks in Fig. 2 are luminance values, and the gray scale of the squares indicates the brightness of the luminance signal. In Figure 2 (a), the brightness of the upper row of gray squares is linearly increasing, that is, increasing from 0.1 to 1.0, and the brightness of the next row of gray squares is incremented according to the power function, that is, the next line of gray The brightness of the square is affected by the distortion of the Gamma nonlinearity. The Gamma characteristic shown by the curve is given in Figure 2 (b).

When multiple links are cascading or connected in series, the total Gamma property is equal to the composition of the Gamma function of each link.

The general model of the Gamma characteristics of multiple links is shown in Appendix S3.

In Figure 3, the nonlinear relationship between the input luminance signal and the output luminance signal at each link is: £out=G ⁽¹⁾ (Zin), Zout=G ⁽²⁾ (in) , out=G ⁽³⁾ (^in

It can be seen that the compound of each link Gamma function is shown as formula (1): G _CT {.) = G ⁽¹⁾ (.) oG ⁽²⁾ (.) oG ⁽³⁾ (.) .·····.. ^ Q oG^Q _(j)

l _0Ul -G _CT (l _in ) = _G ("„ ⁽ "- ²⁾ ( ....... G ⁽²⁾ (G ⁽¹⁾ (/,, ) )))))

Where "." means the compound operation of the function. CT indicates cascaded total, which means the total gamma of the cascade.

In practice, Gamma nonlinearity is caused by different causes. The Gamma characteristics of display devices such as CRT monitors are ideally:

L„ _ut = L _in ^{2 2} (2)

The ideal Gamma characteristics of the corresponding camera/camera are:

From the origin of the Gamma problem, the Gamma problem originated from the CRT display because its Gamma value is 2.2. To compensate for this nonlinearity, a Gamma value of 0.45 was introduced into the camera. In this example, the form of the Gamma property is a Power Function. It should be noted that the input and output luminance signals here are normalized in their respective coordinate spaces, that is, 0 ≤ L. _{ut ≤l, 0≤L} _in _≤l. While other types of displays such as liquid crystals, the form of the Gamma function may be different, or although it is also a power function in form, the parameters are different.

Ideally, there is a linear relationship between the input luminance signal and the output luminance signal, that is, L. _Ut = L _ta . To obtain a linear relationship, Gamma Correction must be performed for links with nonlinear Gamma characteristics.

The principle of Gamma correction for a Gamma link is shown in Figure 4.

In Figure 4, for a link, its Gamma characteristic is given as L. _Ut = Gg(L _in ), so that another correction link 1^1=0 _£ ;(1^) can be cascaded with it to make the total Gamma characteristic a true linear relationship, thus achieving correction The purpose of the nonlinearity of the link.

Obviously, Gg(.) and Gc(.) are inverse functions of each other. In general, for a function, the inverse function is not necessarily solved, and even if there is an inverse function, it cannot be obtained by calculation.

In practical applications, more cases are the case of multiple Gamma links, and the principle of Gamma correction for multiple Gamma links is shown in Fig. 5.

In Figure 5, the correction link needs to be inserted between the two given Gamma links. The Gamma characteristic of the given given link is L. _Ut = Ga(L _in ), the gamma characteristic of a given link is L. _Ut = Gp(L _in ). At this point, the Gc(.) in the correction link is very complicated, and the Gc(.) and Ga(.) or Gp(.) are no longer simple inverse functions.

The above-described Gamma correction method is implemented with the premise that the Gamma characteristic parameter can be determined for a given Gamma link or a cascade of a given Gamma link. The Gamma characteristic parameter here is the parameter of the Gamma characteristic function curve.

In the general case of communication, the correction needs to involve more than two communication terminals. For example, in a two-party video communication, the video of terminal A is transmitted to terminal B, and the correction of the video involves both the Gamma link on terminal A and the Gamma link on terminal B.

The Gamma characteristic parameters need to be determined during the gamma correction process. At present, there are two main methods for determining the Gamma characteristic parameters: Method 1: Instrument measurement method. The input luminance signal and the output luminance signal are measured, that is, some points on the Gamma characteristic function curve are measured by a special instrument, and then the data fitting method is used to perform curve fitting to determine the Gamma characteristic parameter.

Method 2: A method of inputting a luminance signal and outputting a luminance signal. That is, for a given Gamma link, as long as Gg(.) Gc(.) for Gamma correction of Gg(.) can be found by satisfying certain conditions _. For a given Gamma link, as long as Ga(.) and Gp(.) satisfy certain conditions, Ga can be found. .) Gmma-corrected Gc(.) with Gp(.).

It can be seen from the description of the above two methods that the current implementation method for determining the Gamma characteristic parameter has a precondition, that is, the specific value of the input luminance signal and the output luminance signal of the Gamma link that needs to be determined by the Gamma characteristic parameter is clearly known. It is to know all the knowledge of the input luminance signal and the output luminance signal of the Gamma link, that is, the input luminance signal and the output luminance signal can be obtained. Therefore, the above two methods for determining the Gamma characteristic parameters are non-blind measurement methods.

The principle of implementation of the non-blind measurement method is shown in Figure 6.

In Fig. 6, the Gamma characteristic parameter measurement system uses the knowledge of the input luminance signal of the Gamma link and the full knowledge of the output luminance signal to measure the Gamma characteristic parameter. Here, the Gamma link can be a single Gamma link or a cascaded Gamma link.

However, the scope of application of the non-blind measurement method in practical applications is very limited, that is to say, the above preconditions cannot be satisfied in many cases. The following three common applications that cannot satisfy the above preconditions are exemplified below.

Application scenario 1: For streaming media services and applications such as IPTV (Internet Protocol Television), since the program production process has been affected by the Gamma characteristics of the video input device, when the program is broadcast, especially on-demand, etc., It is not possible to obtain the Gamma characteristic of the video input device used to acquire the video signal during the original program production, and it is also impossible to measure the Gamma characteristic parameter of the video input device.

Application Scenario 2: The above problems also exist for applications such as data conferencing. At present, the development of video conferencing and the development of data conferencing are synchronized, and the perfect combination of the two has great significance for collaborative applications. In an enterprise environment, the above-mentioned collaborative application business has strong market demand. However, in data conferencing applications, the source of many multimedia materials such as pictures is untestable, it is difficult to obtain the Gamma characteristics of the video input device that generated the data at the time, and it is also impossible to perform the Gamma characteristic parameters of the video input device. Measured. '

Application Scenario 3: For the public video communication service for millions of home users, in order to reduce the cost and the threshold of video communication service usage, a large number of inexpensive cameras are often used, especially those which are very cheap USB interface cameras. The Gamma characteristic curve and standard L of these inexpensive video input devices. _Ut = Ι^ ⁴⁵ is far apart, not even a power function at all. Moreover, the Gamma characteristic parameters are generally not available from the factory technical data of these inexpensive cameras. Even some cheap cameras simply do not have factory technical information. When the user uses these cameras at home, it is also impossible to obtain the Gamma characteristic parameters by the above method of determining the Gamma characteristic parameters.

As can be seen from the above description, the current non-blind measurement method for determining the Gamma characteristic parameter makes the Gamma correction difficult to apply, and the method of special instrument measurement improves the implementation cost of the video communication service. Summary of the invention

Embodiments of the present invention provide a method and apparatus for correcting gamma characteristics of video communication, which reduces the cost of video communication and improves the ease of use of Gamma correction.

Embodiments of the present invention provide a method for correcting gamma characteristics of a video communication, including:

Obtaining a luminance histogram of the output luminance signal; Converting a luminance histogram of the output luminance signal into a luminance density distribution probability density function of the output luminance signal, and determining an extreme point of the luminance density distribution probability density function of the output luminance signal;

Establishing a mathematical relationship between an extreme point of the luminance density distribution probability density function of the output luminance signal and an extreme point of the luminance density distribution probability density function of the input luminance signal;

Using the extreme value and the mathematical relationship between the extreme points, the mathematical relationship between the luminance density probability density function of the input and output luminance signals and the Gamma characteristic function is converted into: at the extreme point, the luminance signal is output a mathematical relationship between a luminance distribution probability density function and a Gamma characteristic function;

Solving the converted mathematical relationship to determine a gamma characteristic parameter;

Gamma correction is performed on the gamma link according to the gamma characteristic parameter.

An embodiment of the present invention further provides a device for correcting gamma characteristics of a video communication, the device comprising:

Obtaining a histogram module: a luminance histogram for obtaining an output luminance signal, and outputting to the first conversion module;

a first conversion module: configured to convert a luminance histogram of the received output luminance signal into an output luminance signal luminance distribution probability density function, and output to the extreme point calculation module and the second conversion module;

An extreme point calculation module: configured to calculate respective extreme points of the brightness distribution probability density function of the received output luminance signal, and output to the second conversion module;

The storage module: a mathematical relationship between an extreme point of the probability density function of the luminance distribution of the output luminance signal and an extreme point of the luminance density distribution probability density function of the input luminance signal, and a probability density of the luminance distribution of each of the input and output luminance signals The mathematical relationship between the function and the gamma property function;

The second conversion module is configured to: according to the extreme point received, the luminance signal distribution probability density function of the output luminance signal, and the mathematical relationship of the extreme points stored in the storage module, the respective brightness of the input and output luminance signals stored in the storage module The mathematical relationship between the distributed probability density function and the gamma characteristic function is converted to: at the extreme point, the mathematical relationship between the luminance density distribution probability density function of the luminance signal and the Gamma characteristic function is output;

The Gamma characteristic parameter solving module is configured to solve and calculate the converted mathematical relationship to determine a gamma characteristic parameter; and the gamma correction module is configured to perform gamma correction on the gamma link according to the gamma characteristic parameter.

The gamma correction method provided by the embodiment of the present invention only needs to output a histogram of the luminance signal, and thus the gamma parameter determination method of the embodiment of the present invention may be referred to as a full blind measurement method. The method does not require any knowledge of the input of the luminance signal. Therefore, the gamma correction method of the embodiment of the present invention has high application feasibility; thereby improving the gamma correction ease of use and broadening the gamma correction by the technical solution provided by the embodiment of the present invention. The scope of application has improved the user experience and service quality. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic diagram of a model of the link Gamma characteristics;

Figure 2 (a) is a schematic diagram of the Gamma characteristic;

Figure 2 (b) is a schematic diagram 2 of the Gamma characteristic;

Figure 3 is a schematic diagram of a model of gamma cascading with multiple links: Figure 4 is a schematic diagram of the gamma correction principle for a Gamma link;

Figure 5 is a schematic diagram of the Gamma correction principle for a plurality of Gamma links;

6 is a schematic diagram of an implementation principle of a non-blind measurement method in the prior art;

7 is a schematic diagram showing an implementation principle of a method for determining a full-blind Gamma characteristic parameter according to an embodiment of the present invention;

Figure 8 is a diagram showing an example of a luminance histogram of a video signal;

9 is a schematic diagram showing a relationship between an input luminance signal and an output luminance signal luminance distribution probability density function according to an embodiment of the present invention; FIG. 10 (a) is a frame input image in the prior art;

Figure 10 (b) is a frame output image in the prior art;

Figure 10 (c) is a schematic diagram showing the correspondence between the extreme values of the luminance density distribution probability density function of the input luminance signal and the output luminance signal in the prior art;

11 is a schematic diagram showing a probability density function of luminance distribution of an output luminance signal obtained by interpolation and data fitting according to an embodiment of the present invention;

12 is a schematic diagram showing the principle of a brute force search method according to an embodiment of the present invention;

Figure 13 is a schematic diagram of the initial hypercube and its peripheral multilayer hypercube in a two-dimensional case according to an embodiment of the present invention; Figure 14 is a schematic diagram of a brute force search method according to an embodiment of the present invention. Mode for carrying out the invention

In many application scenarios, the specific value of the output luminance signal is known, and the specific value of the input luminance signal is unknown, and even any knowledge of the input luminance signal is unknown. Embodiments of the present invention may utilize only knowledge of the output luminance signal to determine Gamma characteristic parameters and perform Gamma correction. Since any knowledge of the input luminance signal is not utilized in the technical solution of the embodiment of the present invention, the method for determining the Gamma characteristic parameter in the technical solution of the embodiment of the present invention may be referred to as a full blind Gamma characteristic parameter determining method.

The implementation principle of the method for determining the full-blind Gamma characteristic parameter of the embodiment of the present invention is as shown in FIG.

In Fig. 7, for a link in which the Gamma characteristic needs to be determined, the embodiment of the present invention determines the Gamma characteristic parameter of the link based on the knowledge of the known output luminance signal. In the method of determining the full blind Gamma characteristic parameter of the embodiment of the present invention, all knowledge of the output luminance signal is known, but the embodiment of the present invention does not necessarily utilize the entire knowledge of the output luminance signal. The entire knowledge of the output luminance signal is known as the output luminance signal.

After the Gamma characteristic parameters are determined based on the knowledge of the output luminance signal, the Gamma loop can be Gamma corrected according to the Gamma characteristic parameters. The link that needs to determine the Gamma characteristic can be a single given Gamma link, or a cascade combination of multiple given Gamma links.

The method for correcting the gamma characteristic of the video communication provided by the embodiment of the present invention will be described in detail below with reference to the accompanying drawings.

The embodiment of the present invention needs to obtain a luminance histogram of the output luminance signal, and the luminance histogram is as shown in FIG. 8. In Fig. 8, the brightness of the set image is 0 to 255 levels, and the different brightness levels correspond to a brightness distribution probability.

A histogram is a technical term in the field of image processing technology. In fact, a histogram can be a discrete form of distributed probability density function. The video signal is composed of a continuous image of one frame and one frame, and the histogram of the output luminance signal can be obtained from a certain frame image. The method of obtaining a histogram of a luminance signal from an image is a conventional technique and will not be described in detail herein.

The histogram of the output luminance signal can also be performed in other stages, such as obtaining a histogram of the output luminance signal when the output luminance signal is also a one-dimensional signal, and at this time, the output luminance signal is not converted into an image.

There is a close relationship between histogram information and continuous distribution probability density functions. In general, the luminance histogram can be directly obtained from the continuous distribution probability density function; conversely, the continuous distribution probability density function can also be obtained from the luminance histogram by means of data interpolation or fitting. In fact, there is a strict proportional relationship between the histogram information and the continuous distribution probability density function. The above relationship is explained below.

For a cascading combination of a given Gamma link or multiple given Gamma links, the overall set of luminance signals is

{s(t)|teR,0<s(t) <l } , where R represents the entire set of real numbers. That is, the entire set of luminance signals is a set of non-negative time signals whose overall signal amplitude is less than or equal to one. The luminance signal here is a luminance signal in a general sense, so the following description of the relationship between the histogram information and the continuous distribution probability density function is applicable to both the input luminance signal and the output luminance signal. For the sake of simplicity of description, the relationship between the histogram information and the continuous distribution probability density function will be described below by taking the output luminance signal as an example.

Due to random interference, these output luminance signals can be viewed as a random process. The statistical characteristics of these output luminance signals may vary, but the output signals can be classified according to the statistical characteristics of the signals, especially according to the distribution probability characteristics. Any signal as a stochastic process has a distribution probability density function corresponding to it. If the stochastic process is stationary (where the stationary is strictly in the sense of stability), then the distribution probability density function is independent of time; if the stochastic process is not stationary Then, this distribution probability density function may be related to time. Therefore, in general, for a stochastic process s(t) (tER, 0 ≤ s(t) < 1 ), f _s (x, t), te R can be used to represent its distribution probability density function. If it is a stationary stochastic process in a strict sense, f _s (x, t), te R and t are independent, that is, the distribution probability density function does not change with time, in this case, f _s (x, t) = f _s (x) o

The normalization processing of the signal is as follows - if a luminance signal s(t) does not satisfy the condition tE R, o ≤ s(t) ≤ 1, then the luminance signal needs to be normalized to satisfy 1 1 , 0 ≤ 8 (1) ≤ 1. For example: If the actual value range of the signal is [0, SJ, then the normalized signal 3„0) is - (t)=s(t)/ S max (4)

The subscript n in equation (4) means English normalized, meaning normalization.

Correspondingly, if the signal is restored from the normalized value to the actual value, that is, the signal is inverse normalized, the calculation formula is as follows:

According to the definition of the distribution probability density function, the distribution probability density function has the following properties:

£^^,1)(1 = 1, for any 1

And (6)

^ ≥ 0, for any 1

Moreover, for a non-negative signal with a signal amplitude of 1 or less, it satisfies:

f _s (x,t) = 0, x < 0^cx > l (7) That is to say, it is impossible for the signal value to be greater than 1 or less than 0, and the probability is zero.

As a natural inference is:

]^^,1)€^ = 1, for any 1 (8) according to the definition of the probability density function, for a small interval length δ and a range [0, 1] on a point χο, f _s (x ₀ , t) * Prob{x. < s(t) < x ₀ + (9)

Or equivalently f _s (x., t) * Prob{x ₀ - ^δ < s(t) ≤ x. δ} ( 10) where: The symbol Prob represents Probability. The intuitive meaning is that at time t, the luminance signal falls within the interval [x. , x. + ] or [x. - x. The probability of + ^ is approximately equal to f _s (x _a , t) . This is actually a way to turn the continuous distribution probability density function into a discrete probability density. From this, it can be seen that the luminance histogram of the signal can be obtained by such discretization from the continuous probability density function.

For a normalized luminance signal, the [0, 1] interval can be equally divided into N subintervals, each of which has a length of 1/N. The k (k = 0, l, 2, ..., N - l) subintervals are [k N, (k + l) N]. If N is large enough, 1/N is small enough, then you can think of it:

^^(^N ¹ ' ¹ ) - ^Prob ^ N ^{≤ S} W ^≤ ^ N ' k=0,l,2,....,N

N 2 -l ( 11 )

Thus, a sequence of probabilities can be formed:

| A: = 0,l,2...,N - 1} ( 12)

The normalized signal is restored to the unregulated signal space. For example, in video communication, the luminance signal usually takes an integer of 0-255, and a total of 256 levels of brightness. Of course, the luminance signal can also be generalized to 2 ^D -level brightness. In this case, it is necessary to linearly map the unit interval [0, 1] into a set {0, 1, 2, 3, 2° -2, 2° -1}, and each sub-interval is expanded by 2 ^D times. (1/N) 2 ^D . Then the corresponding probability sequence becomes a continuous probability density function -

{h(k) I h(k) = - 1} ( 13)

According to formula (8) and formula (10), it is obvious that:

∑h( = l ( 14) This sequence of probabilities in equation (13) is called the histogram of the luminance signal s(t).

It can be clearly seen from the above derivation that the histogram can be directly obtained from the continuous distribution probability density function of the luminance signal. Conversely, the continuous distribution probability density function of the luminance signal can also be processed by data interpolation, fitting, etc. of the histogram. After getting it.

A specific example of a histogram is given below. When the brightness of the luminance signal includes 256 brightness levels, the correspondence between the probability sequence and the specific values in the histogram is: h(0)=0

h(l)=0 h(64)=0.005

h(65)=0.006 h(190)=0.006

h(l 91 )=0.005

h(192)=0.001

h(193)=0 h(255)=0

In the embodiment of the present invention, the gamma characteristic function may be selected from one of two gamma characteristic functions provided by the following embodiments. Of course, the gamma characteristic function can also be other forms of function, as long as the gamma characteristic function satisfies continuous smoothness and at least second order is achievable.

The following function Υ

To represent the Gamma feature of the Gamma link whose Gamma characteristic parameter is unknown, where the Gamma link includes a single Gamma link or a cascade combination of multiple Gamma links. Above

In the representation of the Gamma property function, P ⁼ [Α, Α, ..., is a parameter vector. In general, the parameter vector consists of 参数 parameters. All or part of these parameters need to be determined. Therefore, according to this very general form, the Gamma property function only needs to satisfy the condition that the function is continuous, and, in general, the Gamma property function is smooth and steerable, at least segmentally smooth and steerable, therefore, assuming Gamma It is reasonable for the characteristic function to exist for the first and second derivatives of the variable X. The first derivative of the Gamma property function can be represented by the following symbol: dx ( 15)

And, the Gamma property function should also satisfy:

g(l;P) = l ( 16)

In general, Gamma property functions can be represented in two common ways:

Method 1, power function:

y = g(x;p) = p,x ^P2 + ρ ₃ ,ρ = [Α, 1⁄2, ⁷ 3 ] ^Γ _{( 1 7} )

Method 2, polynomial function:

y = g(x; p) = Ριχ ^κ + ρ ₂ χ ^κ — ¹ +....+ ρ _κ χ + ρ _κ+1 , where ρ = [ , , ,..., ρ _κ+1 ΐ ( is The above formula (18) can also be transformed into the expression of the formula (19):

y = g(x; p) = Pi (χ-χ.) ^K + ρ ₂ (χ-χ. ) ^{κ 1} +... · + ρ _κ (χ-χ.) + Ρ _κ+ ι ( ¹⁹ )

Where 'p

If e (t) and r (t) are used to represent the input luminance signal and the output luminance signal, respectively, then the distribution probability density functions corresponding to e (t) and r (t) are: f _e (χ, t) and f _r (x, t) , and the following relationship exists between e (t), r (t) and Gamma property functions: r(t) = g(e(t); p), p = [p _i ,p ₂ , p ₃ ,..., p _u f.

According to the probability theory, the relationship as shown in Fig. 9 can be derived, namely:

d(e; p)f _r ( _r ) = f _e (e), where ( ₂₀ )

r = g(e;p),Ve G [0,l]

The specific process of deriving the formula (20) can be referred to the commonly used probability book, which will not be described in detail in this embodiment.

As can be seen from equation (20), equation (20) is independent of time variable t. In fact, the Gamma property function itself is independent of the time variable. Therefore, a set of gamma characteristic parameters are measured over a period of time, and the gamma characteristic parameters can be used throughout the communication during the period of time, such as in IPTV, a program. The Gamma characteristic parameters can be considered to be the same, so that the Gamma characteristic parameters can be measured at the beginning of each program, and the Gamma characteristic parameters can be used throughout the implementation of the program.

After obtaining the luminance density distribution probability density function of the output luminance signal, it is necessary to calculate the respective extreme points of the luminance density distribution probability density function of the output luminance signal. The various extreme points of the calculation function are conventional techniques and will not be described in detail in the embodiments of the present invention.

There is a strict one-to-one correspondence between the extreme point of the input luminance signal luminance distribution probability density function and the extreme value of the output luminance signal luminance distribution probability density function. This correspondence is shown in Figure 10.

In Fig. 10, (a) is a frame of the input video signal, (b) is a frame of the output video signal corresponding to (a), and the two curves in (c) are the brightness of the input luminance signal. Distribution probability density function and output luminance signal luminance distribution probability density function.

As can be seen from Figure (c), the extreme points of the two curves are strictly one-to-one correspondence. This correspondence is:

Er^ (maximum point), e ₂ ->r ₂ (minimum point), e ₃ ->r ₃ (maximum point), e ₄ ->r ₄ (minimum point), e ₅ ->r ₅ (maximum point).

Without loss of generality, the input luminance signal brightness distribution probability density function and the output luminance signal luminance distribution probability density function have J extreme points, respectively e, e ₂ , ...., and n, r ₂ Rj.

Based on the correspondence of the extreme points, the extreme point e _{k of the} luminance density distribution probability density function of the input luminance signal and the extreme value point r _{k of the} corresponding luminance signal distribution probability density function (k=l, 2,...,J) A reasonable approximation is performed, that is, the setting formula (21) is established:

r _k = g(e _k ;p), P = [A, A, ,···,/1⁄2 k=l,2,3 .,J (21 )

Where: k = l, 2, 3, ..., J; J is the number of extreme points of the luminance distribution probability density function.

After slightly changing the formula (21), the process of determining the gamma parameter is basically the same as the following description, and will not be described in detail here.

It can be seen from the above description that the extreme point of the luminance density distribution probability density function of the input luminance signal and the extreme value point of the luminance density distribution probability density function of the output luminance signal have the following two relationships: relationship 1, qualitative geometric topological relationship, ie two There is a one-to-one correspondence between the extreme points of the luminance density probability density function, that is, the number of extreme points of the two luminance distribution probability density functions is the same. Relationship 2, quantitative mathematical relationship: rk gp^ p - Q^;^;^...,;^] . Of course, the mathematical relationship here allows A little change.

It is set that f _e (x) and f _f (x) are continuously steerable, and d(x; p) is continuously steerable, that is, g(x; p) is second-order continuous steerable. In this way, you can derive the two sides of the formula (20), you can get:

Z _(e;P) f _{r +} d ² _(e;P) ^ = ^, where

Dr de

r = g(e;p), (22)

Since _ei, e _2, ej is f _e (x) of the extreme point, therefore, in these extreme point derivative f _e (x) is zero. Thus, the following formula exists:

3⁄4^| _e _ _e =0,k = l,2,3"..,J (23)

De ^k

Combining equation (22) with equation (23), the following relationship holds: z(e _k ;p)f _r (r) + d ² (e _k ;p)^ = 0 ₍₂₄₎

r = g(e _k ;p),k = l,2,3,....,J

Since the formula (21) is approximately true, the formula (24) can be transformed into the following relationship: z(e _k ; p)f _r (r _k ) + d ² (e _k ; p) = 0, k = 1,2, 3,...., J (25)

Dr z(g ^_1 (r _k ; P); P)f _r (r _k ) + d ² (g- ¹ (r _k ; p); p) = 0,k = 1,2,3,... ., J (26)

Dr

In equation (26), k = l, 2, 3, J, J is the number of extreme points of the luminance distribution probability density function, r _k is the extreme point of the luminance function of the output luminance signal, and the derivative function is:

Dr

^ ^ = dc^- ¹ + (d- l)c _d , r ^d - ² +(d- 2)c _d . ₂ r ^d - ³ +..... + c, .

Dr

In practical applications, the Gamma property function is monotonic, so the Gamma property function has an inverse function. This inverse function can be written as g_ '(χ;ρ). Obviously, the inverse of the Gamma property function is also dependent on the gamma parameter vector p.

It can be seen from equation (26) that equation (26) does not contain any information about the input signal. Equation (26) is completely dependent on the output luminance signal brightness distribution probability density function and the output luminance signal luminance distribution probability density function. Value point. Thus, embodiments of the present invention are capable of determining Gamma characteristic parameters based solely on the output signal and its luminance distribution probability density function.

The histogram of the output luminance signal obtained by setting the embodiment of the present invention is:

.., Nl}. That is, each histogram contains N items, and in the histogram terminology, each item is called a "bin".

According to the formula (12), the luminance density distribution probability density function of the output luminance signal can be obtained from the output image luminance histogram, and thus:

?k + 1

+ i _{= Nh (k)} ^ - ₌ 0,1,2..., N - 1 (27)

' 2N ^r for N points uniformly distributed over the interval [0, 1] a _k = ^~5", 1^ = 1, 2, ...., :^ can get the function)

The value on the 2N scatter, ie f _f (^^) A: = 0, l, 2..., N - l. The position of these discrete points on the r-axis of the coordinate system is as shown in FIG.

2N

If N is large enough, the expression of the probability density of the output luminance distribution can be obtained by interpolation or data fitting: f _r (r) = c _d r ^d + c _d ., r ^d - ¹ + c _d . ₂ r ^d - ² +―. + c,r + c ₀ (28)

Where: c _d , c _d-1 , c _d-2 , ... c _c is d+1 polynomial coefficients, and r is the amplitude of the output luminance signal.

This expression is a polynomial including a polynomial spline function.

In general, for an image of 256 levels of brightness, N = 256, at which point N is already large enough. There are various implementation methods of interpolation or data fitting techniques, and embodiments of the present invention do not limit the specific implementation of interpolation or data fitting.

The derivative of equation (28) is:

= dcZ ¹ + (d - l)c _d - - ² + (d - 2)c _d . ₂ r ^d - ³ +...-. + _Cl (29)

Dr

The present invention can select the expression of the gamma characteristic function of any one of the expressions given in (17), (18), (19). At this time, the specific form of the Gamma characteristic function is obtained, but only the parameter vector p needs determine. Calculate any of the equations (26) for each given value of the parameter vector p, that is, calculate each multiplication factor and addition term in equation (26).

If there are J equations, as long as J ≥ M-1, where M is the number of parameters that need to determine the gamma characteristic, then the only solution can be solved by this set of equations as the M-1 gamma parameters. The value is determined, and then the constant term is determined by using equation (16) as a constraint, that is, determining the remaining one gamma parameter.

That is to say, due to the existence of the formula (16), all the gamma parameters satisfy a constraint condition. If these gamma parameters are M, then only M-1 parameters are independent, as long as it is determined Any M-1 parameters of the M gamma parameters, and the remaining gamma parameter can be solved by the formula (16), thereby determining all the parameters of the Gamma characteristic function.

In general, the condition J ≥ M-1 can be satisfied.

There are various methods for obtaining the parameter vector p according to the formula (26). The following mainly introduces two methods for determining the parameter vector p: Method 1. Solve the equation directly.

J equations are obtained from equation (26), and M-1 equations are arbitrarily selected to form a simultaneous solution of equations. In general, this set of equations is non-linear and transcendental, such as the Gamma property function in the form of a power function. Therefore there is no closed-form solution. A numerical solution method is required. The numerical solution of the system of equations is a conventional technique and will not be described in detail in this embodiment.

Method two, nonlinear function optimization method.

Construct a cost function according to formula (26):

Obviously, for the true parameter vector p _{trae of the} Gamma property function, it should theoretically make J(p _tnie )=0, since the parameter vector Ptrue satisfies each equation in equation (30), therefore, in equation (30) Each summation is zero. However, in practical applications, because there are errors and approximations such as approximations in obtaining the histogram process, each summation term in equation (30) will not be equal to zero, but it should be a small value, and The following conditions should be met: For any _P ER ^M , there is J(p J(p _true ). That is, p, _rae is the global minimal point of the function J(p).

As can be seen from the above description, for any of the Gamma characteristic functions given by equations (17), (18), and (19), g- can be calculated based on each given value of the parameter vector p. i(r _k ;p), zCg-'Cr^p^p) , d ² (g ^_1 (r _k ;p); p), meanwhile, the luminance density distribution probability density function of the output luminance signal is known, such that You can follow formula (28),

^f _r (r _k )

(29) Calculate ^f _r ( ^f k) and dr so that each summation term in equation (30) can be calculated, and the total summation result J can be calculated.

Therefore, in the nonlinear function optimization method, the problem of determining the parameter vector p is transformed into a mathematical problem of finding the global minimum point of the cost function J(p).

In the nonlinear function optimization method, the parameter vector p can be determined in the following three ways.

Mode (1), conventional mathematical optimization method.

Since there are derivatives in J(p), classical mathematical optimization methods such as gradient method, conjugate gradient method, etc. can be used to determine the parameter to fip.

The specific process of determining the parameter vector P by the conventional mathematical optimization method is a conventional technique and will not be described in detail in this embodiment.

Mode (2), neural network method.

The specific process of determining the parameter vector p by the neural network method is a conventional technique and will not be described in detail in this embodiment. Method (3), Brutal Force Search Method. The so-called brute force search, as the name suggests, is to search for all possibilities.

For the case where the parameter takes a discrete value, then the set of all possible values of the parameter is a finite set. Then, by searching each point in the set one by one, the point that minimizes J(p) can be found, and thus the global minimum is found. Point p _true . However, this situation is rare. In most cases, the parameters take continuous values. Therefore, the set of all possible values of the parameters is an infinite set, and the exhaustive search cannot be performed.

For the case where the parameter takes a continuous value, the brute force search method can divide the parameter space into a plurality of small hypercubes, and then take a point in each hypercube as a sample point, such as the geometric center point of the hypercube. Etc. Finally, calculate the function value of the cost function at the sampling point of each hypercube, find the sampling point of the hypercube that minimizes the cost function, and use the sampling point as the global minimum point.

The implementation process of determining the parameters of the Gamma characteristic function by the brute force search method will be described in detail below with reference to the case where the parameters take continuous values.

In the embodiment of the present invention, a prior knowledge of the parameters can be utilized to find a suitable starting search point, so that Large reductions in the number of searches required, thereby increasing the efficiency of the brute force search method.

The principle of the brute force search method is shown in Figure 12.

In Fig. 12, the set of all possible values of the M-1 parameters constitutes a parameter space (PS), and the parameter space is a subset of the M-1 dimensional European space RM- ¹ . After the parameter space is determined, the implementation method of the brute force search includes the following steps: Step 1: Hypercube partitioning.

The PS is divided into a plurality of M-1 hypercubes, in Fig. 12, ABCDEFGH, 8 points form a hypercube. Since each parameter has a different range of values, each side of the hypercube has a different length.

Set the k (k=l, 2 M-1) parameter p _{k to} the range of [min _k , maxj, P _k aliquots the kth dimension, so that the edge of each hypercube in this dimension Long is

Thus the volume of each hypercube is:

Therefore, the total number of hypercubes is greater than Τ - Π^Ρ^. This value is the maximum possible. If the shape of the PS is not a hypercube, then the number of long cubes it contains may be much smaller than T = π^-'ρ.

Use the indicator vector for each hypercube! : , ,...,^—^ to represent, where i _k (k=l, 2, ..., M-1 , i _k =l , 2,

3 P _k ) indicates that in the kth dimension, the hypercube is the i _kth .

So, for the first! For the two ^^...^^^ hypercubes, the coordinate range in each dimension is:

[min _k + (i _k - l)A _k , min _k + i _k A _k ] (33)

The geometric center of the hypercube is <3⁄4 coordinates -

[min! + Ί - Ι^Δ,, πώ^ + ( ₂ - 1/2) Δ ₂ , , ιηΐη^ + ( _Μ ., - 1/2) Δ _Μ .,] ^Τ (34) Step 2, Select Initial search point.

In general, for a parameter Pk, there is a reasonable value, which can be used as the initial value. For example, when the Gamma characteristic function is in the form of a power function, the value of the Gamma parameter is generally around 2.2 for the video input device. Because the industry standard requires 2.2, the Gamma parameter may be positive or negative due to manufacturing technology and product quality. Deviated from 2.2, but in most cases, it is closer to 2.2. In this case, if the search is started with 2.2 as the initial search point, since 2.2 is closer to the true value, the number of attempts to find the real value is less. By the same token, each parameter p _k (k=l, 2, ..., Ml) can be found with its appropriate initial value p _intk , then these initial values form a vector, which is the initial value of the parameter vector p _Pint = [p _intll , _Pintk , Pin,, p _int - n ] Step 3, Start the search based on the selected initial search point.

First, determine which hypercube Pint falls in, and compare the coordinates and other methods to determine which hypercube _{Pint is} in. The specific implementation process of the comparison coordinate method is:

Set the coordinates of the hypercube to be I _tat

, the judgment condition is: For k=0, 1, 2..... M-1, If and only if the following formula (35) holds, it is determined that p _int falls in the hypercube with the coordinates: ^^ [^,^,...^^(^. min _k + (i _Mk - l)A _k < p _intk < min _k + ϊ^Δ, ( 35 )

After determining the hypercube in which the initial search point is located, the search starts from the hypercube. Calculating J ( _Pint ) according to the formula (30), if J (Pi „ _t ) can make the formula (36 ), or the searched hypercube satisfies a predetermined condition, such as the number of searched hypercubes reaches a predetermined value, etc., then , the entire search process ends, to step 5. At this point, _Ptrue

The advantage is the parameter vector of the finally obtained Gamma property function.

J ( Pint) ≤ Jthreshold ( 36)

Among them, Jthre _S h. Ld is a predefined threshold.

If J ( _Pint ) does not make equation (36) true, go to step four.

Step 4. Continue searching.

The continuation search in step four can be done hierarchically. The hypercube surrounding the outside of the initial hypercube can be one or more layers.

In the two-dimensional case, the initial hypercube and its peripheral multi-layer hypercube are shown in Figure 13.

In Fig. 13, the middle gray square is the initial hypercube, the cube adjacent to the edge of the initial hypercube is the first layer hypercube, the cube adjacent to the edge of the first layer hypercube is the second layer hypercube, and the second The cheap cube on the side of the layer hypercube is the third layer of hypercube.

The hierarchical search method of the embodiment of the present invention is: successively searching each of the hypercubes in each layer except the initial hypercube, and in each layer of the search for the hypercube, each of the layers should be traversed in a predetermined order. Hypercube. The predetermined order may be varied, and the embodiment of the present invention does not limit the form of the predetermined order as long as it can traverse the hypercube in one layer.

In a layer of hypercube search, when searching for a hypercube according to a predetermined order, the coordinates of the geometric center Q of the hypercube need to be calculated according to formula (34), and then the function value J (Q) is calculated. J ( Q) can make formula (37), or the number of searched hypercubes reaches a predetermined value, then the entire search process ends, to step 5. At this point, Ptrue

= Q;

J ( Q ≤ Jthreshold ( 37 )

If J ( Q) does not enable equation (37 ) and the number of searched hypercubes does not reach the predetermined value, continue searching in this layer of hypercubes according to the predetermined order. If the hypercube search for this layer is completed and the K Q) of each supercube in this layer is unable to make equation (37) true, continue searching for the next layer of hypercube.

When the L-th layer is searched, the search process will end regardless of whether or not the geometric center Q of the hypercube that satisfies the condition (37) is found. At this time, the Q in the smallest J (Q) searched should be taken as Ptrue. Go to step five. Here, L is a predetermined threshold value indicating the number of layers that need to be searched at most.

Step five, the search process ends.

The above search method can also be applied to achieve the fastest and best search effect in the process from coarse to fine.

The search process from coarse to fine is as shown in Fig. 14.

In Fig. 14, first, the first search is performed in accordance with the above steps 1 through 5. The first search can be seen as a rough search, this In this way, you can set the side length of the hypercube in the parameter space to be larger, so that the number of hypercubes in the parameter space is small. In the search process from coarse to fine, the predetermined condition that the hypercube searched in step 3 satisfies may be whether the division granularity is coarser than the predetermined division granularity. First search If a hypercube that satisfies condition (36) or (37) is found, the search process ends. The first search process can use a hierarchical search method.

If the hypercube that satisfies condition (36) or (37) is not searched during the first search, and the granularity of the hypercube has not reached the predetermined granularity, the smallest J (Q) found in the first time. The second finer search is performed in the corresponding hypercube. At this time, the smallest J (Q) corresponding hypercube searched for the first time should be regarded as the new entire parameter space. At this time, the length of each hypercube in the new parameter space becomes smaller, and the search process from the first step to the fifth step is repeated. Similarly, in the second, thinner search process, if a hypercube that satisfies condition (36) or (37) is found, the second search process ends. The second search process can employ a hierarchical search method.

If the hypercube that satisfies the condition (36) or (37) is not found in the second finer search process, and the granularity of the hypercube has not reached the predetermined granularity, the smallest one found in the second search. The third finer search is performed in the corresponding hypercube of J (Q), and so on, the geometric center of the hypercube that satisfies the condition (36) or (37) found by the last fine search is taken as the global best advantage Pttue, That is, the parameter vector of the Gamma property function.

If the granularity of the hypercube reaches the predetermined granularity, but the hyperelement satisfying the condition (36) or (37) is still not found, the search process ends. At this point, the collection center of the hypercube corresponding to the smallest J (Q) found should be taken as the global best advantage Ptrue, the parameter vector of the SPGamma property function.

A hierarchical search method can be employed in each of the above-described search processes from coarse to fine.

After the parameter vector of the Gamma characteristic function is determined by the above method, the Gamma link can be corrected by Gamma. Here, the link that needs to perform the gamma correction may be a video data source device, an intermediate device in the video communication network, or a video data destination device.

As can be seen from the description of the above technical solution, the gamma correction method provided by the embodiment of the present invention only needs to know the histogram of the output luminance signal and a set of loose assumptions to determine the gamma characteristic parameter of a given gamma link, thereby The multimedia communication system provides an easy-to-implement gamma correction method. The gamma correction method provided by the embodiment of the present invention has high application feasibility, thereby greatly broadening the application range of gamma correction, especially for IPTV, collaborative data conference, Public video communication using low-end video input devices provides a good gamma correction function, greatly improving user experience and service quality, further enhancing the competitiveness of these services, and bringing huge benefits to telecom operators, service providers and equipment manufacturers. Economic benefits.

The apparatus for correcting video communication gamma characteristics provided by the embodiments of the present invention mainly includes: acquiring a histogram module, a first conversion module, an extreme point calculation module, a storage module, a second conversion module, a Gamma characteristic parameter solving module, and a gamma correction Module.

The acquisition histogram module is mainly used to obtain a luminance histogram of the output luminance signal, and output the obtained luminance histogram to the first conversion module. The acquisition histogram module can obtain a luminance histogram of the output luminance signal before converting the output luminance signal into an output image frame, and can also obtain a luminance histogram of the output luminance signal from the output image frame. Specifically, it is described in the above method.

The first conversion module is mainly configured to convert a luminance histogram of the received output luminance signal into a luminance density distribution probability density function of the output luminance signal, and output the luminance signal distribution probability density function of the output luminance signal to the extreme point calculation module and the second conversion respectively. Module. Here the output luminance signal brightness distribution probability density function can be - f _r (r) = c _d r ^d + Cd - - ¹ + c _d . ₂ r ^d " ² + -.... + c]r + c ₀ .

The extreme point calculation module is mainly configured to calculate each extreme point of the brightness distribution probability density function of the output brightness signal after receiving the brightness distribution probability density function of the output brightness signal, and transmit the calculated extreme points to the second conversion Module. The method of calculating the extreme point of the luminance density distribution probability density function of the luminance signal is a conventional technique and will not be described in detail herein.

The storage module is mainly used for storing the mathematical relationship between the extreme point of the luminance density distribution probability density function of the output luminance signal and the extreme point of the luminance density distribution probability density function of the input luminance signal, and storing the luminance density probability density of each of the input and output luminance signals. The mathematical relationship between the function and the gamma property function is d(e;p)f _f (r) = f _e (e).

Here, the mathematical relationship between the extreme point of the brightness distribution probability density function of the output luminance signal and the extreme point of the luminance density distribution probability density function of the input luminance signal may be: rk gie ppf;^/^, ^,...,/ ^]^ Of course, the mathematical relationship here allows for a slight transformation, as described in the above method.

The second conversion module is mainly used for respectively distributing the brightness distribution of the input and output luminance signals in the storage module according to the extreme points received, the luminance density distribution probability density function of the output luminance signal, and the mathematical relationship of the extreme points stored in the storage module. The mathematical relationship d(e;p)ff (r) = f _e (e) between the probability density function and the gamma characteristic function is converted into: the probability density function of the luminance distribution of the output luminance signal and its derivative function at the extreme point The mathematical relationship between the Gamma characteristic function and its inverse, first derivative, and second derivative: z(g - ¹ (r _k ;p); p)f _r (r _k ) ₊ d ² (g- ¹ (r _k ;p);p)^3⁄4) = ₀ ,k = 1,2,3,....,J;

Dr

Where: k = l , 2, 3 , J, J is the number of extreme points of the luminance distribution probability density function, r _k is the extreme point of the luminance distribution probability function of the output luminance signal, and the derivative function ^ ^ is:

Dr

= ¹ + (d - l)c _d , r ^d - ² + (d - 2)c _d . ₂ r ^d - ³ + ..... + _Cl .

Dr

The Gamma characteristic parameter solving module is mainly used to solve the mathematical relationship after the conversion of the second conversion module to determine the gamma characteristic parameter. The Gamma characteristic parameter solving module can determine the gamma characteristic parameters by directly solving the equation method, the nonlinear function optimization method, and the like. The nonlinear function optimization methods herein include conventional mathematical optimization methods, neural network methods, brute force search, and the like. The Gamma characteristic parameter solving module can adopt a hierarchical search method when using the brute force search method, or a hierarchical search method from coarse to fine, as described in the above method.

The gamma correction module is mainly used for gamma correction of the gamma link according to the Gamma characteristic parameter obtained by the Gamma characteristic parameter solving module. The gamma link here includes: a cascading combination of a given Gamma link or multiple given Gamma links.

The device provided by the embodiment of the present invention is located in a video device, such as in a video data source device, in an intermediate device of a video communication network, and in a video data destination device.

While the invention has been described by the embodiments of the invention, it will be understood that

Claims

Rights request

A method for correcting gamma characteristics of a video communication, comprising:

Obtaining a luminance histogram of the output luminance signal;

Converting a luminance histogram of the output luminance signal into a luminance density distribution probability density function of the output luminance signal, and determining an extreme point of the luminance density distribution probability density function of the output luminance signal;

2. The method of claim 1 wherein the step of converting the luminance histogram of the output luminance signal to the output luminance signal luminance distribution probability density function comprises:

Converting the luminance histogram of the output luminance signal into a polynomial form of the output luminance signal luminance distribution probability density function: f _r (r) = c _d r ^d + Cd- - ¹ + c _d . ₂ r ^d - ² +...-. + c]r + c ₀ ;

Where: c _d , c _{d-1 >} c _d-2 , c _Q is d+1 polynomial coefficients, and r is the amplitude of the output luminance signal.

3. The method of claim 2, wherein:

The mathematical relationship between the luminance distribution probability density function f _e (x, t;), f _r (x, t) , and the Gamma characteristic function of each of the input and output luminance signals is set as follows:

d(e;p)f _r (r) = f _e (e) _;

Where: r = g(e,p), Ve e [0,l] _;

Set the maximum number of extreme points of the luminance distribution probability density function of the input and output luminance signals, and the quantitative mathematical relationship is:

Rk = g(e _k ; P), P = [A , A, A,... , ΡΜΪ

Where: k = l , 2, 3, ..., J; J is the number of extreme points of the luminance distribution probability density function;

At the extreme point, the mathematical relationship between the converted luminance signal density distribution density function and its derivative function and the Gamma characteristic function and its inverse and derivative functions is: z(g- ¹ (r _k ;p) ;p)f _r (r _k ) + d^g-'ir^p^p)^^ = 0,k = 1,2,3,...., J;

Or

Where: k = l, 2, 3, J, J is the number of extreme points of the luminance distribution probability density function, r _k is the extreme point of the luminance distribution probability function of the output luminance signal, and the derivative function is:

Dr = dc _d r ^d -' + (d - l)c _d , r ^d - ² + (d - 2)c _d . ₂ r ^d - ³ +..... + c, .

Dr

4. The method according to claim 3, wherein the step of solving the converted mathematical relationship to determine a gamma characteristic parameter comprises:

Solve the mathematical relationship by using the direct solution equation method or the nonlinear function optimization method: z(g ^_1 (r _k ;P); P)f _r (rk) + 1,2,3,„..,J Determine the gamma characteristic parameters.

5. The method of claim 4, wherein the step of determining a gamma characteristic parameter using a nonlinear function optimization method comprises:

Constructing a cost function according to the mathematical relationship between the brightness distribution probability density function of the output luminance signal and the Gamma characteristic function: ' J(P) = ;

Selecting M-1 gamma characteristic parameters arbitrarily from M gamma characteristic parameters, and reducing the dimension of the gamma characteristic parameter space to M-1 dimension;

The method for determining the M-1 gamma characteristic parameters is: determining the gamma characteristic parameter vector _Ptrue so that for any pe R ^M - ¹ , the relationship J

(P)

That is for this

The value determined by the M-1 gamma characteristic parameters;

Use the relationship g ^(1; P) ^{= 1 to} combine

, determine the remaining gamma characteristic parameter.

6. The method of claim 5, wherein the method of determining F e comprises: a combined optimization method, or a neural network method, or a brute force search method.

7. The method of claim 6, wherein the brute force search method comprises the steps of:

Dividing the dimensionally reduced one-dimensional gamma characteristic parameter space into multiple hypercubes;

Selecting an initial search point, and performing a traversal search from the hypercube in which the initial search point is located according to a predetermined order; calculating a geometric center coordinate Q of each hypercube that enters during the search, and entering according to the cost function and the search process Hypercube calculation J (Q);

If J (Q) is less than or equal to the predetermined threshold, or the searched hypercube satisfies the predetermined condition, the geometric center coordinate Q of the hypercube entered in this search is taken as _Ρ(ηκ) .

8. The method of claim 7, wherein the initial search point is set according to an empirical value of a gamma characteristic parameter in an actual application.

9. The method of claim 7, wherein the traversal search comprises: a hypercube in which an initial search point is located

1 X As the initial hypercube, the hypercube surrounding the initial hypercube is divided into a multi-layered hypercube array that sequentially wraps the previous layer according to the distance from the initial hypercube, and searches layer by layer.

10. The method of claim 7 wherein:

In the brute force search method, the dimensionally reduced one-dimensional gamma characteristic parameter space is divided into a plurality of coarse-grained hypercubes, and the traversal search includes:

Taking the hypercube in which the initial search point is located as the initial hypercube, the hypercube surrounding the initial hypercube is divided into multi-layer hypercube arrays which are sequentially wrapped in the previous layer according to the distance from the initial hypercube, and searched layer by layer;

The searched hypercube that satisfies the condition is used as a new gamma characteristic parameter space, and is divided into finer-grained hypercubes, and so on, and a layer-by-layer search from coarse to fine.

11. A device for correcting gamma characteristics of a video communication, the device comprising:

12. Apparatus according to claim 11 wherein said apparatus is located in a video data source device, and/or in an intermediate device of a video communication network, and/or is located in a video data destination device.

1 Q