CN101272488B - Video decoding method and device for reducing LCD display movement fuzz - Google Patents

Abstract

The invention discloses a video decoding method for reducing the LCD displaying motion fuzzy, which includes the following steps: a video stream enters a de-multiplexing module to be de-multiplexed to obtain motion coding information and texture coding information; after the motion coding information enters a motion coding module, the vector information of motion estimation is analyzed, and the vector information is input into a motion compensation module to obtain a motion compensation image; the texture coding information enters an image texture decoding module to obtain a texture decoding image which is combined with the motion compensation image to obtain a reconstructed image; the reconstructed image is input into an image marginal detecting module to obtain marginal vector information, which enters a dynamic image marginal enhanced module together with the motion estimation vector information; the dynamic image marginal enhanced module inputs enhanced video images into an LCD displaying module. The invention also discloses a device for realizing the method. The method of the invention can effectively improve the motion fuzzy phenomenon during the LCD displaying and save software and hardware resources as well as reduce implementation costs.

Description

Reduce the video encoding/decoding method and the device of LCD display movement fuzz

Technical field

The present invention relates to a kind of video encoding/decoding method, relate in particular to the video encoding/decoding method and the equipment therefor thereof of a kind of reduction LCD (LiquidCrystal Display, liquid crystal) display movement fuzz.

Background technology

Than traditional C RT (Cathode Ray Tube, cathode ray tube) display, a subject matter of LCD display is motion blur, promptly can cause fuzzy or conditions of streaking when showing the video scene of rapid movement.

Cause the main cause of LCD display movement fuzz to comprise:

(1) response time is slow: the response time of liquid crystal can be caused the Scene Blur of Fast transforms slowly, studies show that this problem accounts for all and causes 30% of motion blur reason.Its solution is to improve the response time of liquid crystal or adopt overload to handle in display driver.

(2) retention effect: different with the moment luminescence mechanism of traditional C RT, LCD display adopts the display mechanism of sampling-maintenance (sample-and-hold), and is promptly all luminous in the whole demonstration time period.And human eye vision has the low-pass filtering effect on time shaft.So cause the motion blur phenomenon.Studies show that this problem accounts for all and causes 70% of motion blur reason.Its solution can be to adopt " scanning backlight " to simulate the CRT display mode.Another kind method is input video to be carried out " moving image enhancing " handle.

The basic thought that " moving image enhancing " handled is video image to be carried out dynamic edge strengthen, and causes the low-pass filtering effect of motion blur with compensation.Strengthen differently with the common still image edge of image processing field, the dynamic edge enhanced strength in " moving image enhancings " processing and scope and this edge strength, edge direction and movement velocity, the direction of motion are relevant.For example, if edge direction perpendicular to the direction of motion, its motion blur effects maximum is so the edge strengthens intensity and scope all answer maximum; Opposite if edge direction is parallel to the direction of motion, its motion blur is zero, also should be zero so the edge strengthens intensity.

" moving image enhancing " handled needs following processing module:

(1) Image Edge-Detection module is used for detected image edge strength, edge direction;

(2) interframe movement estimation module is used to detect movement velocity, the direction of motion;

(3) the dynamic image edge strengthens.

Handle in the required module in above-mentioned " moving image enhancing ", the interframe movement estimation module is calculated the most complicated, and required software and hardware resources is more, is not suitable for the embedded system real-time implementation.

Summary of the invention

The technical problem to be solved in the present invention provides a kind of video encoding/decoding method of the LCD of reduction display movement fuzz, can overcome the retention effect problem that causes the LCD display movement fuzz, and replace the interframe movement estimation module with existing motion estimation vectors in the conventional video decoding, save software and hardware resources, reduce and realize cost.For this reason, the present invention also will provide a kind of video decoder of realizing this method.

In order to solve the problems of the technologies described above, the video encoding/decoding method that the present invention reduces the LCD display movement fuzz comprises the steps:

(1) Shu Ru video flowing enters demultiplexing module and carries out demultiplexing, to obtain motion encoded information and texture coding information;

(2) the motion encoded information of step (1) gained enters the motion estimation vectors information that solves each image block in the image behind the motion decoder module, this motion estimation vectors information input motion compensating module, and in this motion compensating module, combine with the reconstructed image of former frame or multiframe, obtain motion compensated image;

(3) the texture coding information of step (1) gained enters the image texture decoder module and obtains the texture decoder image, and this texture decoder image combines in image reconstruction module with the motion compensated image of step (2) gained, obtains the reconstructed image of this frame;

(4) reconstructed image input picture rim detection module obtains edge vectors information, motion estimation vectors information through motion decoder module decoding gained in this edge vectors information and the step (2) enters dynamic image edge enhancing module as input parameter, makes this dynamic image edge strengthen module inputted video image is carried out the enhancing of dynamic edge;

(5) the dynamic image edge strengthens module the video image input LCD display module demonstration after strengthening.

A kind of video decoder that reduces the LCD display movement fuzz, comprise demultiplexing module, motion decoder module, image texture decoder module, motion compensating module, and image reconstruction module, also comprise: Image Edge-Detection module and dynamic image edge strengthen module

Described demultiplexing module obtains motion encoded information and texture coding information according to the video flowing of input, and described motion encoded information is exported to the motion decoder module, and described texture coding information is exported to the image texture decoder module;

Described motion decoder module obtains the motion estimation vectors information of each image block in the image according to described motion encoded information, exports to motion compensating module and dynamic image edge again and strengthens module;

Described motion compensating module obtains motion compensated image according to described motion estimation vectors information and the former frame of combining image rebuilding module input or the reconstructed image of multiframe;

Described image texture decoder module obtains the texture decoder image according to described texture coding information;

Described motion compensated image and texture decoder image are exported to image reconstruction module together, and described image reconstruction module obtains the reconstructed image of this frame, export to motion compensating module and Image Edge-Detection module again;

Described Image Edge-Detection module and image reconstruction module, dynamic image edge strengthen module and are connected, and are used to receive the reconstructed image of described image reconstruction module input, and the edge vectors message transmission that will detect gained is given dynamic image edge enhancing module;

Described dynamic image edge strengthens module and is connected with Image Edge-Detection module, motion decoder module and LCD display module, be used to receive the edge vectors information of described Image Edge-Detection module input and the motion estimation vectors information of described motion decoder module input, and determine the edge enhanced strength according to this edge vectors information and motion estimation vectors information, video image is carried out being transferred to the LCD display module after dynamic edge strengthens.

The video encoding/decoding method of reduction LCD display movement fuzz of the present invention, motion blur phenomenon when showing by conventional video decoding process and LCD moving image enhancing flow process being combined closely, effectively improving LCD, and significantly reduce amount of calculation, simplify handling process, reduce and realize cost.The inventive method is fit to Implementation of Embedded System, and particularly mobile multi-medium system is realized, as multimedia handset, PDA etc.

Description of drawings

Below in conjunction with the drawings and specific embodiments the present invention is described in further detail.

Accompanying drawing is the flow chart that the present invention reduces the video encoding/decoding method of LCD display movement fuzz.

Embodiment

The video encoding/decoding method of reduction LCD display movement fuzz of the present invention is that the conventional video decoding is combined with LCD moving image enhancement techniques, utilize the motion estimation vectors information in the conventional video encoding stream to combine with the edge vectors information that Image Edge-Detection obtains, obtain the dynamic image edge and strengthen desired parameters, inputted video image is carried out dynamic edge strengthen, the motion blur phenomenon was effectively improved when it was shown on LCD.

As shown in drawings, video encoding/decoding method of the present invention comprises the steps:

(1) Shu Ru video flowing enters demultiplexing module and carries out demultiplexing, to obtain motion encoded information and texture coding information;

(2) the motion encoded information of step (1) gained enters the motion estimation vectors information that solves each image block in the image behind the motion decoder module, this motion estimation vectors information input motion compensating module, and in this motion compensating module, combine with the reconstructed image of former frame or multiframe, obtain motion compensated image;

(3) the texture coding information of step (1) gained enters the image texture decoder module and obtains the texture decoder image, and this texture decoder image combines in image reconstruction module with the motion compensated image of step (2) gained, obtains the reconstructed image of this frame;

(4) reconstructed image input picture rim detection module obtains edge vectors information, motion estimation vectors information through motion decoder module decoding gained in this edge vectors information and the step (2) enters dynamic image edge enhancing module as input parameter, makes this dynamic image edge strengthen module inputted video image is carried out the enhancing of dynamic edge;

(5) the dynamic image edge strengthens module the video image input LCD display module demonstration after strengthening.

Wherein, demultiplexing module, motion decoder module, the image texture decoder module, motion compensating module, and image reconstruction module is all identical with conventional video decoding framework, its implementation is because different video standard and different, as MPEG2, MPEG4, H.264, video standard such as AVS.

The Image Edge-Detection module has a lot of application in digital image processing field.The method for detecting image edge that many maturations are arranged at present, for example can adopt Sobel operator commonly used to carry out rim detection in the method for the invention:

Vertical operator: $S_v=\left[\begin{array}{ccc}-1& -2& -1\\ 0& 0& 0\\ 1& 2& 1\end{array}\right]$ Horizontal operator: $S_h=\left[\begin{array}{ccc}-1& 0& 1\\ -2& 0& 2\\ -1& 0& 1\end{array}\right]$

Earlier with horizontal operator and vertical operator image being carried out convolution respectively, obtain two matrixes, also is image M 1, the M2 onesize with original image under the situation of not considering the border, and they distinguish two partial derivatives at same position place among the presentation video M.Then addition after two numbers of M1, M2 correspondence position square is obtained a new matrix G, G represents the Grad of the gray scale of each pixel among the M, thereby obtains the edge vectors direction and the intensity of each pixel in the image.Though the difference as a result that different edge detection methods obtain, the present invention is not limited to use certain specific edge detection method.

It is to decide the edge enhanced strength according to the edge vectors direction of each pixel in the image and the direction and the intensity of intensity and motion estimation vectors that the dynamic image edge strengthens module.The angle of edge vectors direction and motion vector direction (0～90 degree) is big more, and the edge enhanced strength also should be big more.The intensity of edge vectors is big more, the intensity of motion vector is big more, and then the edge enhanced strength also should be big more.According to this principle, the enhancing function is at method definable dynamic image of the present invention edge:

$D(i,j)=\frac{\left|{\mathrm{EV}}_{i,j}\right|}{{\left|\mathrm{EV}\right|}_{\max }}\×\frac{\left|{\mathrm{MV}}_{i,j}\right|}{{\left|\mathrm{MV}\right|}_{\max }}\×\sin (\∠{\mathrm{EV}}_{i,j},{\mathrm{MV}}_{i,j})\×f\left(I_{i,j}\right)$

Wherein | EV _{I, j}| be the edge vectors intensity of this point, | EV| _MaxBe edge vectors intensity maximum in the image; | MV _{I, j}| be the motion vector intensity of this point, | MV| _MaxBe edge vectors intensity maximum in the image; ∠ EV _{I, j}, MV _{I, j}Represent the edge vectors of this point and the angle of motion vector (0～90 degree).F (I _{I, j}) represent that image is done the edge strengthens filtering, for example available following Laplce's factor pair image is done convolutional filtering and is strengthened the edge:

$f=\left[\begin{array}{ccc}-1& -1& -1\\ -1& 9& -1\\ -1& -1& -1\end{array}\right]$

The present invention is not limited to use the dynamic image edge of above definition to strengthen function and edge enhancing filter function.Other implementation methods that strengthen under the principle at aforesaid dynamic image edge also are applicable to the present invention.

The inventive method is not only applicable to improve the LCD display effect, is applicable to the display effect that improves other non-stroboscopic display unit simultaneously, as plasma scope, and OLED (OrganicLight-Emitting Diode, Organic Light Emitting Diode) display etc.

The video decoder of reduction LCD display movement fuzz of the present invention, on the basis of conventional video decode block shelf structure, increase Image Edge-Detection module and dynamic image edge and strengthened module, described Image Edge-Detection module strengthens module with image reconstruction module, dynamic image edge and is connected, be used to receive the reconstructed image of described image reconstruction module input, and the edge vectors message transmission that will detect gained is given dynamic image edge enhancing module; Described dynamic image edge strengthens module and is connected with Image Edge-Detection module, motion decoder module and LCD display module, be used to receive the edge vectors information of described Image Edge-Detection module input and the motion estimation vectors information of described motion decoder module input, and determine the edge enhanced strength according to this edge vectors information and motion estimation vectors information, video image is carried out being transferred to the LCD display module after dynamic edge strengthens.

Video decoder of the present invention is applicable to most video standards, as MPEG2, MPEG4, H.264, AVS etc., all be suitable for device of the present invention so long as comprise the video standard of motion estimation vectors information in the interframe encode information.

Claims (2)

1. a video encoding/decoding method that reduces the LCD display movement fuzz is characterized in that, comprises the steps:

(1) Shu Ru video flowing enters demultiplexing module and carries out demultiplexing, to obtain motion encoded information and texture coding information;

(2) the motion encoded information of step (1) gained enters the motion estimation vectors information that solves each image block in the image behind the motion decoder module, this motion estimation vectors information input motion compensating module, and in this motion compensating module, combine with the reconstructed image of former frame or multiframe, obtain motion compensated image;

(3) the texture coding information of step (1) gained enters the image texture decoder module and obtains the texture decoder image, and this texture decoder image combines in image reconstruction module with the motion compensated image of step (2) gained, obtains the reconstructed image of this frame;

(4) reconstructed image input picture rim detection module obtains edge vectors information, motion estimation vectors information through motion decoder module decoding gained in this edge vectors information and the step (2) enters dynamic image edge enhancing module as input parameter, makes this dynamic image edge strengthen module inputted video image is carried out the enhancing of dynamic edge;

(5) the dynamic image edge strengthens module the video image input LCD display module demonstration after strengthening.

2. video decoder that reduces the LCD display movement fuzz, comprise demultiplexing module, motion decoder module, image texture decoder module, motion compensating module, and image reconstruction module, it is characterized in that, also comprise: Image Edge-Detection module and dynamic image edge strengthen module

Described demultiplexing module obtains motion encoded information and texture coding information according to the video flowing of input, and described motion encoded information is exported to the motion decoder module, and described texture coding information is exported to the image texture decoder module;

Described motion decoder module obtains the motion estimation vectors information of each image block in the image according to described motion encoded information, exports to motion compensating module and dynamic image edge again and strengthens module;

Described motion compensating module obtains motion compensated image according to described motion estimation vectors information and the former frame of combining image rebuilding module input or the reconstructed image of multiframe;

Described image texture decoder module obtains the texture decoder image according to described texture coding information;

Described motion compensated image and texture decoder image are exported to image reconstruction module together, and described image reconstruction module obtains the reconstructed image of this frame, export to motion compensating module and Image Edge-Detection module again;

Described Image Edge-Detection module and image reconstruction module, dynamic image edge strengthen module and are connected, and are used to receive the reconstructed image of described image reconstruction module input, and the edge vectors message transmission that will detect gained is given dynamic image edge enhancing module;

Described dynamic image edge strengthens module and is connected with Image Edge-Detection module, motion decoder module and LCD display module, be used to receive the edge vectors information of described Image Edge-Detection module input and the motion estimation vectors information of described motion decoder module input, and determine the edge enhanced strength according to this edge vectors information and motion estimation vectors information, video image is carried out being transferred to the LCD display module after dynamic edge strengthens.

Images