KR20200079871A - Image processing method and display device using the same - Google Patents

Abstract

The present invention relates to an image processing method for improving accuracy in an occlusion area, and a display device using the same. The image processing method synthesizes at least one of an IMG1 which is a first image in an n^th frame, an MC1 which is a forward motion compensation image using the IMG1, an IMG2 which is a second image in an (n+1)^th frame, and an MC2 which is a backward motion compensation image using the IMG2, to generate an interpolated image in an interpolation frame.

Description

Image processing method and display device using the same}

The present invention relates to an image processing method and a display device using the same.

The liquid crystal display is driven by a hold type display. Due to the driving characteristics of such a hold type display, motion blur may be observed in a video, resulting in a problem that image quality deteriorates.

To improve the motion blur problem, there are a scanning backlight and a black insertion technique. This method expresses a black image by temporarily turning off the backlight for a period of 1 frame. When the eye of the person who remembers the previous image looks at the next screen, the previously memorized image is a black image and does not appear as an afterimage. This method is very effective and easy to implement, but has a disadvantage in that power efficiency is relatively deteriorated due to the need to emit higher light in a short period of time.

As another method for improving the motion blur problem, a frame rate up-conversion (FRUC) technique is in the spotlight. FRUC is a technique of additionally inserting frames between frames of an input original image.

The algorithms of FRUC can be broadly classified into two types, one that does not consider the motion of an object, and the other that considers the motion of an object. In the former case, it is simple to implement in the same way as frame linear interpolation or frame repetition, but it has disadvantages such as blurring and motion jerkiness for objects with motion. This happens.

On the other hand, in the latter case called motion compensation FRUC (Motion Compensated Frame Rate Up-Conversion, MC-FRUC), blurring or blurring occurs because interpolated frames are generated in consideration of the motion of objects in the image. FRUC can be performed without motion jerkiness. This method uses a lot of computation and a lot of frame memory because, like when compressing an image, the edges of the real image (uncompressed) must be extracted and recognized as a vector, and frame data must be stored and new frames created and sent at a high frequency. IC (Integrated Circuit) is required. Therefore, this method has the disadvantage that it is difficult to implement and the IC price is high, but the power efficiency of the backlight is very high because the brightness of the screen is uniform.

The motion compensation FRUC (MC-FRUC) can be largely divided into motion estimation (ME) and motion compensated (MC). Motion determination (ME) is a process of calculating the motion of an object between successive frames to obtain a motion vector (MV) corresponding to a change in motion, and motion compensation (MC) is performed through motion determination (ME) This is a process of generating a new frame image for interpolation based on the obtained motion vector (MV).

Although various algorithms such as a Pel-recusive algorithm, a phase correlation algorithm, and a block matching method are known as the motion determination (ME), a block matching algorithm (BMA) is generally used due to convenience of implementation.

Motion compensation (MC) is a one-way motion compensation method (uni-D MC) that generates a motion compensation image (MCI) using a motion vector (MV) and one original image, and a motion vector (MV) and two or more originals There is a two-way motion compensation method (bi-D MC) that generates an image motion compensation image (MCI). The bidirectional motion compensation method has an advantage in that the quality of the background region is superior to the unidirectional compensation method. This is because it is difficult to obtain hidden background image information through a one-way compensation method.

The motion compensation image (MCI) generated by the motion compensation (MC) is not always accurate. In some cases, an inaccurate motion compensation image (MCI) may be generated. When an inaccurate motion compensation image (MCI) is generated as described above, the FRUC method that generates an interpolated image image by repetition of a frame may generate a better image quality. To solve this problem, the motion compensation FRUC (MC-FRUC) determines the accuracy of the motion compensation image (MCI), and determines the method of synthesizing the motion compensation and the original image of the input image according to the accuracy to obtain an image of the interpolated frame. Will print.

According to a method of determining a motion compensation and a method of synthesizing the original image of the input image, it is classified into a motion vector (MV) based method and a data based method. Since the data-based method uses high resolution data, it is disadvantageous in terms of cost and requires high technology. However, it has the advantage of expecting excellent performance when securing accuracy due to high resolution.

The above-mentioned background technology is technical information acquired by the inventor for the derivation of the present invention or acquired in the derivation process of the present invention, and is not necessarily a known technology disclosed to the general public prior to the filing of the present invention.

An object of the present invention is to improve the accuracy in the occlusion region by generating an interpolation image using the correlation between the motion compensation and the original image of the input image, and finally, to improve the image quality in the region.

As a means for solving the above-described problems, the present invention has an embodiment with the following features.

The present invention includes IMG1 as the first image in the n frame, MC1 as the forward motion compensation image using the IMG1, IMG2 as the second image in the n+1 frame, and MC2 as the backward motion compensation image using the IMG2. It is characterized by generating at least one of the interpolated image in the interpolation frame.

The image processing method includes: an MC synthesis step of synthesizing the MC1 and the MC2 to calculate an MC synthesis value MC_S; An original image synthesis step of synthesizing the IMG1 and the IMG2 to calculate an original image synthesis value IMG_S; And a final synthesis step of synthesizing the MC composite value MC_S and the original image composite value IMG_S to generate an interpolated image IMG_out. It characterized in that it comprises.

을 통하여 MC 합성 값 MC_S를 산출하는 것을 특징으로 한다.The MC synthesis step, Equation

It is characterized by calculating the MC composite value MC_S through.

The forward motion compensation image MC1 includes the uncovered area in which the background obscured by the foreground in the IMG1 is revealed by the movement of the foreground or background in the IMG2 due to the overlapping of the background and the background in the IMG1. If it is, the wc is characterized by having a value of 0.

When the forward motion compensation image MC1 has no overlap between the foreground and the background in the IMG1, but the overlap between the foreground and the background occurs in the IMG2 due to the movement of the foreground or the background, and the background includes a cover area that is obscured by the foreground. , Wc has a value of 1.

The original image synthesis step,

을 통하여 원본 이미지 합성 값 IMG_S를 산출하는 것을 특징으로 한다.Equation

It is characterized by calculating the original image synthesis value IMG_S through.

The wi is characterized by dividing the frequency of the input image by the frequency of the output image in which the frame is increased due to the generation of an interpolation frame.

을 통하여 보간 영상 IMG_out을 생성하는 것을 특징으로 한다.The final synthesis step, Equation

It is characterized by generating the interpolation image IMG_out through.

The wf increases as the difference between MC1 and MC2 increases; Smaller differences between IMG1 and IMG2 increase; A smaller wocc indicating the degree to which the IMG1 and the IMG2 include occlusion regions having a change in overlap between the foreground and the background due to the relative movement of the foreground or background within the IMG1 or the IMG2; It is characterized by.

The degree of inclusion of the occlusion region calculates mu, which is an average value of MC1, MC2, IMG1, and IMG2; D_mu, which is a difference value of each of the mu and the MC1, MC2, IMG1, and IMG2 is calculated; The images of which D_mu is the largest among MC1, MC2, IMG1, and IMG2 are classified as D_max, and the remaining images are classified as D_else; Calculate mu_else, which is an average value of the images classified as D_else; Calculating a difference value D_alone between the images classified as mu and D_max; Calculate by calculating D_else, which is a difference value between mu and mu_else; The wocc is characterized by being a difference value between the D_alone and D_else.

In addition, the present invention, IMG1 as the first image in the n frame, MC1 as the forward motion compensation image using the IMG1, IMG2 as the second image in the n+1 frame, and backward motion compensation using the IMG2 An image processor configured to generate an interpolated image in the interpolation frame by synthesizing at least one of the images MC2; And it characterized in that it comprises a display panel for displaying an image based on the image output from the image processing unit.

The image processing unit includes: an MC synthesizer that synthesizes the MC1 and the MC2 to calculate an MC composite value MC_S; An original image synthesizer which synthesizes the IMG1 and the IMG2 to calculate an original image synthesis value IMG_S; And a final synthesizer combining the MC composite value MC_S and the original image composite value IMG_S to generate an interpolated image IMG_out. It characterized in that it comprises.

In the present invention, in generating an interpolation image by using a correlation between the motion compensation and the original image of the input image, motion synthesis is determined, synthesis of the original image is determined, and finally motion compensation and synthesis of the original image are determined. By determining and performing the interpolation image step by step, there is an effect of generating an interpolation image with high accuracy while minimizing cost increase.

In addition, the present invention generates interpolation in the occlusion area by generating an interpolation image using correlation between forward motion compensation and backward motion compensation images, two original video images, and a total of four images in order to consider the influence of the occlusion area. It has the effect of high image accuracy.

1 is a schematic block diagram of a display device
2 is a schematic circuit diagram of a sub-pixel
3 is a view for explaining the concept of FRUC
4 is a conceptual diagram for obtaining a motion vector
5 is a view for explaining a unidirectional motion compensation method
6 is a view for explaining a bi-directional motion compensation method
7 is a diagram for describing an intermediate motion vector field
8 is a block diagram of an image processing unit according to an embodiment of the present invention
9 is a block diagram of an image synthesis module according to an embodiment of the present invention
10 is a view for explaining the concept of image synthesis
11 is a diagram for determining a wc value of a foreground region in an MC synthesis method.
12 is a diagram for determining a wc value of an uncover region in the MC synthesis method
13 is a diagram for determining a wc value of a cover region in the MC synthesis method
14 is a view for explaining an element for determining wocc in the occlusion region.
15 is a flowchart of a method for calculating wocc

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic block diagram of a display device, and FIG. 2 is a schematic circuit diagram of a sub-pixel.

The display device according to an exemplary embodiment of the present invention includes a display panel 100, a gate driver 200, a source driver 300, a timing controller 400, and an image processing unit 500.

The display panel 100 displays an image corresponding to the data signal DATA and the scan signal supplied from the source driver 300 and the gate driver 200. The display panel 100 includes sub-pixels SP that operate to display an image.

The sub-pixels SP include a red sub-pixel, a green sub-pixel and a blue sub-pixel, or a white sub-pixel, a red sub-pixel, a green sub-pixel and a blue sub-pixel. The display panel 100 may be divided into a liquid crystal display panel, an organic light emitting display panel, and an electrophoretic display panel according to the configuration of the pixel circuit PC included in the subpixels SP.

As shown in FIG. 2, one sub-pixel SP is supplied corresponding to the scan signal supplied through the switching transistor SW and the switching transistor SW connected to the scan line GL1 and the data line DL1. It includes a pixel circuit (PC) that operates in response to the data signal. When the display panel is selected as the organic light emitting display panel, the pixel circuit PC of the sub-pixel SP may be configured such that various types of compensation circuits are further added together with a driving transistor, a storage capacitor, and an organic light emitting diode.

The gate driver 200 outputs a scan signal while shifting the level of the gate voltage in response to the gate timing control signal GCS supplied from the timing controller 400. The gate driver 200 outputs a scan signal through scan lines GL1 to GLm. The gate driver 200 may be formed in the form of an integrated circuit (IC) or a gate in panel method on the display panel 100.

The source driver 300 samples and latches the data signal DATA supplied from the timing controller 400 in response to the data timing control signal DCS supplied from the timing controller 400, converts it into a gamma reference voltage, and outputs it. do. The source driver 300 outputs the data signal DATA through the data lines DL1 to DLn. The source driver 300 may be formed in the form of an integrated circuit (IC).

The timing controller 400 includes a data signal (including a driving signal including a data enable signal DE or a vertical synchronization signal Vsync, a horizontal synchronization signal Hsync and a clock CLK signal) from the image processing unit 500. DATA). The timing controller 400 includes a gate timing control signal GCS for controlling the operation timing of the gate driver 200 based on a driving signal and a data timing control signal DCS for controlling the operation timing of the source driver 300. Output

The image processing unit 500 outputs a data enable signal DE and the like as well as a data signal DATA supplied from the outside. The image processing unit 500 may output one or more of a vertical synchronization signal, a horizontal synchronization signal, and a clock signal in addition to the data enable signal DE. In addition, the image processing unit 500 includes a FRUC-capable module for improving image quality based on information about an input image.

As a method to improve the motion blur problem, there is a frame rate up-conversion (FRUC) technique.

FRUC's algorithms can be broadly classified into two types, one that does not consider the motion of an object. This method is simple to implement in a method such as frame linear interpolation or frame repetition, but has a disadvantage that blurring occurs for an object having motion.

3 is a diagram for explaining the concept of FRUC.

Another method is motion compensation FRUC, which generates an interpolated frame in consideration of the motion of objects in an image. It is common to consider the motion of objects in block units. As illustrated in FIG. 4, n+ is the first block B_n included in the n-th frame (frame [n]) and the second block B_n+1 which is a block matching the first block B_n. By searching in the first frame (frame [n+1]), the motion of the block can be grasped.

The process of calculating the motion of an object between successive frames to obtain a motion vector (MV) corresponding to a change in motion is called motion determination (ME). Also, a process of generating a new frame image for interpolation based on a motion vector (MV) obtained through motion determination (ME) is called motion compensation (MC). As in the case of compressing an image, this method requires extracting the edge of a real image (uncompressed) and recognize it as a vector. It is a method of storing frame data and creating a new frame and sending it at a high frequency. This method has the advantage that FRUC can be performed without blurring or motion jerkiness.

The image processing unit 500 of the present invention performs FRUC in a motion compensation manner. Hereinafter, the motion compensation FRUC will be simply referred to as FRUC.

4 is a conceptual diagram for obtaining a motion vector. The process of obtaining the motion vector MV will be described in detail with reference to FIG. 5.

Motion determination (ME) generally has various methods such as a block matching method, a 3-D recursive search method, a hierarchical search method, and a MAP application method. In the present invention, an embodiment is constructed using a method based on block matching, but is not limited thereto. The method used in the block matching can be m x n pixel as one motion vector block unit. In frame [n], the block containing the individual object to measure motion is B_n, and in frame [n+1], the block matching B_n is B_n+1. In each block, motion vectors are stored in the form of horizontal and vertical values of the motion of the image from frame [n] to frame [n+1]. That is, the motion vector can be expressed in the form of MV=(mv_x, mv_y).

Block matching is a process of obtaining image information in frame [n] (brightness, RGB, etc.) and image information in frame [n+1] at the location with the lowest cost. Cost is generally sum of absolute Difference (SAD) is the most widely used. The SAD value refers to the sum of absolute values of the difference in image information of all pixels included in a block.

A block matching value based on frame [n] can be expressed as MV1, and a value obtained based on frame [n+1] as MV2. MV1 and MV2 have the same size and have opposite directions or opposite signs. In general, the motion determination ME may include all motion vectors MV obtained based on a plurality of frames, including frame [n] and [n+1].

5 is a diagram for explaining a unidirectional motion compensation method, FIG. 6 is a diagram for explaining a bidirectional motion compensation method, and FIG. 7 is a diagram for explaining an intermediate motion vector field.

When the motion compensation (MC) is described in detail, the motion compensation (MC) is the first image IMG1 or the n+1 frame (frame [n+1) of the nth frame (frame [n]) together with the motion determination information. ]) is the process of generating the NX frame (frame [nx]) (hereinafter interpolated frame) using the second image IMG2 or a plurality of image information including the same. In addition, the rectangular box image generated in the N.X frame (n.x) (hereinafter interpolated frame) by the compensation operation at this time is referred to as a motion compensated image (MCI).

As illustrated in FIG. 5, in the unidirectional motion compensation method, motion compensation is performed using only one of the first image IMG1 or the second image IMG2. It is called "uni-directional MC" because only one-way video is used.

As illustrated in FIG. 6, the bi-directional motion compensation method performs motion compensation using a plurality of images including a first image IMG1 and a second image IMG2. It is called "bi-directional MC" because it uses both images.

The process of making motion compensation using the first image (IMG1) is called forward motion compensation (FW-MC), and the process of making motion compensation using the second picture (IMG2) is backward motion compensation (backward) MC; BW-MC).

As illustrated in FIG. 7, the unidirectional and bidirectional motion compensation schemes perform interpolation frame (frame [nx]) to perform one or more of forward motion compensation (FW-MC) and backward motion compensation (BW-MC). The direction of the corresponding motion vector is aligned in the forward (+) or reverse (-) direction. At this time, a set of motion vectors obtained by scaling the size of an interpolation frame (frame [n.x]) is called an intermediate motion vector field (IMVF).

Since the uni-directional motion compensation method (Uni-D MC) uses only one image selected from the first image IMG1 or the second image IMG2, the degree of freedom in allocating the buffer memory is high. However, this method is not free from image quality deterioration such as motion artifacts when implementing a video.

The two-way motion compensation method (Bi-D MC) is relatively free from deterioration of image quality such as MotionArtifact because both images of the first image IMG1 and the second image IMG2 are used. However, this method requires twice as much buffer memory as the one-way motion compensation method.

8 is a block diagram of an image processing unit according to an embodiment of the present invention.

The image processing unit 500 of the present invention performs FRUC in a motion compensation manner. The image processing unit 500 includes a motion vector calculation module 510, a compensation image generation module 520, and an image synthesis module 530.

The motion vector calculation module 510 calculates a motion vector (MV) from an image input including a plurality of images. The process of calculating the motion of an object between successive frames of an input image to obtain a motion vector (MV) corresponding to a change in motion is called motion determination (ME). The motion vector (MV) is used to generate a compensation image.

The compensation image generation module 520 may include the first image IMG1 or the n+1 frame (frame [n+1]) of the n-th frame (frame [n]) together with the motion vector MV that is the motion determination information. ), MCI, which is an image of an interpolation frame, is generated using the second image IMG2 or a plurality of image information including the same.

The image synthesis module 530 determines the accuracy of the MCI, synthesizes the original image input to the MCI and the image processing unit 500 according to the accuracy, and outputs the final image of the interpolation frame. The synthesis method is based on a motion vector (MV) calculated by the motion vector calculation module 510, and a method of using a correlation between an original video image and MCI.

9 is a block diagram of an image synthesis module according to an embodiment of the present invention, and FIG. 10 is a diagram for explaining the concept of image synthesis. If the image synthesis module 530 is described with reference to FIGS. 9 to 10,

The video synthesis module 530 includes IMG1 as a first image in n frames, MC1 as a forward motion compensation image using the IMG1, IMG2 as a second image in an n+1 frame, and backward motion compensation using the IMG2 At least one of the images MC2 is synthesized to generate an interpolated image in the interpolation frame.

In FIG. 10, IMG1 is an original image in a t frame, IMG2 is an original image in a t+1 frame, and F_i is an interpolation frame.

The present invention combines the original image IMG1 in the t frame, the image IMG2 in the t+1 frame, the MC1 which is the forward motion compensation image, and the MC2 which is the backward motion compensation image to make the image of the interpolation frame. Produces

The interpolation image IMG_out generated by the image synthesis module 530 may be expressed by Equation 1 below.

The characteristic of the weight parameter w for each image is equal to wc1+wc2+wi1+wi2=1.

Since the complexity increases when the four weighting parameters w are calculated at the same time, if Equation 1 is rearranged in order to obtain this step by step, it can be expressed as Equation 2 below.

If Equation 2 is categorized and sorted by item again, it can be expressed as Equations 3 to 5.

The image synthesis module 530 includes an MC synthesizer 531, an original image synthesizer 532, and a final synthesizer 533. The image synthesis module 530 may generate an interpolation image with high accuracy while minimizing cost increase by performing the interpolation image generation step by step.

The MC synthesizer 531 calculates the MC composite value MC_S through Equation (4).

11 is a diagram for determining a wc value of a foreground region in the MC synthesis method. Referring to FIG. 11, a method of determining wc when the information of the image is the foreground (FG) or the pure background (BG) will be described.

The method of determining wc may be any one of values from 0 to 1 when the information of the image is the foreground (FG) or the pure background (BG).

In FIG. 11, when the gray image in the t frame (F_t) moves to the right from the t+1 frame (F_t+1), the original image of the t+1 frame (F_t+1) is used by using a motion vector (MV). Alternatively, the MC image may be created by referring to the original image of the t frame (F_t). Forward MC (FW-MC) refers to the process of creating an MC1 video with reference to the original video of t+1 frame (F_t+1), and backward MC refers to the process of creating an MC2 video with reference to the original video of t frame (F_t). It is called (BW-MC). When the motion vector MV is correct, the composite MC image MC_S generated in the interpolation image frame F_i is the same in the foreground FG with reference to any image of MC1 and MC2.

Also, in the case of a pure background (BG), it can be described in the same way as the foreground (FG). Therefore, in theory, when the information of the image is the foreground (FG) or the pure background (BG), the value of wc is the same regardless of the value from 0 to 1.

In this case, the pure background refers to a case where the frame is changed from t to t+1 and remains covered or uncovered by the foreground.

12 is a diagram for determining a wc value of an uncover region in the MC synthesis method, and FIG. 13 is a diagram for determining a wc value of a cover region in the MC synthesis method.

The cover is not a pure background, that is, a hidden background is called an uncover. (The sum of cover and uncover is called occlusion.) In the uncover region, wc is 0, and in the cover region, wc is 1. In the present invention, a filtering method using a motion vector (MV) is applied to the uncover/cover region when determining wc, and is omitted because it is not the main content of the present invention.

In FIG. 12, the background (BG) image is in a stationary state, and the foreground (FG) image is in a state of moving to the right. An uncover region, which is a background region exposed due to the movement of the foreground (FG) image, is generated. In the uncover area, MC1 cannot obtain accurate image information because the background (BG) is covered by the foreground (FG). Therefore, in this case, it is desirable to set the wc value to 0 and exclude the MC1 image from the synthesis of MC.

That is, in the forward motion compensation image MC1, the background is obscured by the overlapping of the foreground and background in the IMG1, and in the IMG2, an uncover region in which the background obscured by the foreground in the IMG1 is revealed by the movement of the foreground or background. When included, wc has a value of 0.

In FIG. 13, the background (BG) image is in a stationary state, and the foreground (FG) image is in a state of moving to the left. A cover area, which is a background area obscured by the movement of the foreground (FG) image, is generated. In the cover area, MC2 cannot obtain accurate image information because the background (BG) is covered by the foreground (FG). Therefore, in this case, it is desirable to set the wc value to 1 and exclude the MC2 image from the synthesis of MC.

That is, in the forward motion compensation image MC1, there was no overlap between the foreground and the background in the IMG1, but in the IMG2, the overlap between the foreground and the background occurs due to the movement of the foreground or the background, and the cover region covers the foreground. In case, wc has a value of 1.

The original image synthesizer 532 calculates the IMG synthesis value IMG_S through Equation 5.

The method of determining wi may be determined differently according to the region characteristics or may be determined according to input/output frequencies. In the present invention, a method of determining according to the input/output frequency was applied. That is, wi may be determined by dividing the frequency of the input image by the frequency of the output image in which the frame is increased due to the generation of the interpolation frame.

For example, in the case of 30 Hz input and 60 Hz output, the interpolation frame to be generated corresponds to 0.5 time-step. In this case, wi is always equal to 0.5.

When two or more generation frames appear, it may be changed to a value between 0 and 1 according to the time-step of the generation frame.

The final synthesizer 533 synthesizes the MC synthesized value MC_S and the original image synthesized value IMG_S through Equation 3 to generate an interpolated image IMG_out. wf has a value from 0 to 1.

The method for determining wf is as follows.

First, wf increases as the difference between MC1 and MC2 increases. Because, if the difference between MC1 and MC2 is large, the accuracy of the MC image is low, so the specific gravity of MC_S is lowered to synthesize.

Second, wf increases as the difference between IMG1 and IMG2 is small. This is because if the difference between IMG1 and IMG2 is small, the MC also has a higher probability of being similar to the original video, so it is okay to synthesize it by increasing the weight of IMG_S, which is an image of the original video.

Third, wf increases as the wocc indicating the degree to which the IMG1 and the IMG2 include occlusion regions having a change in overlap between the foreground and the background due to the relative movement of the foreground or background within the IMG1 or the IMG2 is increased as the wocc is smaller. . This is because the difference between MC1 and MC2 increases in the occlusion area, so that the effect on this is offset.

If the expression of wf is reflected by the above-described determination method, it can be expressed as in Equation (6).

c1, c2, c3, and c4 are constants, and the sign can be both negative and positive. For example, c1, c2, c3, and c4 may have values of 1, -1, -1, and 0 in order.

14 is a view for explaining an element for determining wocc in the occlusion region, and FIG. 15 is a flowchart of a method for calculating wocc. When explaining a method for determining wocc with reference to FIGS. 14 to 15,

14 shows a still gray foreground image and a white background image moving to the left. Because of the occlusion region, the images of IMG1, IMG2, and MC1 are all the same in white, whereas only MC2 is gray. When the wOcc value is determined using these characteristics, wocc can be determined to reflect a case where any one of the four images of MC1, MC2, IMG1, and IMG2 has a large difference.

Referring to Figure 15 in detail,

1) The average values of MC1, MC2, IMG1, and IMG2 mu are calculated (S10).

2) D_mu, which is a difference value of each of the mu and the MC1, MC2, IMG1, and IMG2 is calculated (S20).

3) Among the MC1, MC2, IMG1, and IMG2, the image having the largest D_mu is classified as D_max, and the other images are classified as D_else (S30).

4) Calculate mu_else, which is the average value of the images classified as D_else (S40).

5) D_alone, which is a difference value between the images classified as mu and D_max, is calculated (S50).

6) Calculate by calculating D_else, which is a difference value between mu and mu_else (S60).

7) The wocc is a difference value between the D_alone and D_else (S70).

wocc = |D_alone-D_else|

It should be understood that the embodiments described above are illustrative in all respects and not restrictive. The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts should be interpreted as being included in the claims of the present invention. .

100: display panel
200: gate driver
300: source driver
400: timing controller
500: image processing unit
510: motion vector calculation module
520: compensation image generation module
530: video synthesis module
531: MC synthesizer
532: Original image synthesizer
533: final synthesizer

Claims (14)

In the image processing method for generating an interpolation frame between n frames and n + 1 frame,
At least one of IMG1 as the first image in the n-frame, MC1 as a forward motion compensation image using the IMG1, IMG2 as a second image in the n+1 frame, and MC2 as a backward motion compensation image using the IMG2 An image processing method for generating an interpolated image in the interpolation frame by synthesizing one. According to claim 1,
The image processing method,
An MC synthesis step of synthesizing the MC1 and the MC2 to calculate an MC synthesis value MC_S;
An original image synthesis step of synthesizing the IMG1 and the IMG2 to calculate an original image synthesis value IMG_S; And
A final synthesis step of synthesizing the MC composite value MC_S and the original image composite value IMG_S to generate an interpolated image IMG_out; Image processing method comprising a. According to claim 2,
The MC synthesis step,
An image processing method comprising calculating the MC composite value MC_S through the following equation.

The wc has a value from 0 to 1. According to claim 3,
The forward motion compensation image MC1 includes an uncover area in IMG1 where the background that is obscured by the foreground is revealed in the IMG1 by the foreground or the movement of the background in the IMG2 due to the overlap of the foreground and the background. If, the wc has a value of 0, characterized in that the image processing method. According to claim 3,
When the forward motion compensation image MC1 has no overlap of the foreground and background in the IMG1, but the overlap of the foreground and the background occurs in the IMG2 due to the movement of the foreground or background, and the background includes a cover area that is obscured by the foreground. , Wherein wc has a value of 1. According to claim 2,
The original image synthesis step,
An image processing method characterized by calculating an original image synthesis value IMG_S through the following equation.

The wi has a value from 0 to 1. The method of claim 6,
The wi is an image processing method characterized by dividing the frequency of the input image by the frequency of the output image whose frame is increased due to the generation of an interpolation frame. According to claim 2,
The final synthesis step,
An image processing method characterized by generating an interpolated image IMG_out through the following equation.

The wf has a value from 0 to 1. The method of claim 8,
The wf is,
Image processing method characterized in that the larger the difference between MC1 and MC2 increases. The method of claim 8,
The wf is,
Image processing method characterized in that the smaller the difference between IMG1 and IMG2 increases. The method of claim 8,
The smaller the wocc indicating the degree to which the IMG1 and the IMG2 include occlusion regions having a change in overlap between the foreground and the background due to the relative movement of the foreground or background within the IMG1 or the IMG2, the wf increases Image processing method characterized in that. The method of claim 11
The extent of occlusion includes
Calculate mu, which is the average value of MC1, MC2, IMG1, and IMG2,
D_mu, which is a difference value of each of the mu and the MC1, MC2, IMG1, and IMG2, is calculated,
Among MC1, MC2, IMG1, and IMG2, the image with the largest D_mu is classified as D_max, and the other images are classified as D_else.
Calculate mu_else, which is the average value of the images classified as D_else,
D_alone, which is a difference value between the images classified as mu and D_max, is calculated,
Calculate by calculating D_else, which is a difference value between mu and mu_else,
The wocc is a difference value between the D_alone and D_else. At least one of IMG1 as the first image in the n-frame, MC1 as a forward motion compensation image using the IMG1, IMG2 as a second image in the n+1 frame, and MC2 as a backward motion compensation image using the IMG2 An image processing unit that synthesizes one to generate an interpolated image in the interpolation frame; And
A display device including a display panel for displaying an image based on the image output from the image processing unit. The method of claim 13,
The image processing unit,
An MC synthesizer that synthesizes the MC1 and the MC2 to calculate an MC synthesis value MC_S;
An original image synthesizer which synthesizes the IMG1 and the IMG2 to calculate an original image synthesis value IMG_S; And
A final synthesizer combining the MC composite value MC_S and the original image composite value IMG_S to generate an interpolated image IMG_out; Display device comprising a.

Images