KR100618635B1 - 3d dither algorithm - Google Patents

Abstract

Four-time resolution refinement of the three-dimensional dither algorithm is provided in the present invention. In the present invention, 4x2 blocks including two 2x2 blocks are treated as an observed unit. In order to eliminate the movement of the lines and the dithering of the edges, at least two LSBs of pixels are processed on a case-by-case basis. For the first 2x2 block, when the 2-bit LSB is 01 or 11, the pixels assigned to carry carry the top-left, bottom-right, bottom-left, and within 2x2 blocks for four consecutive frames. It will be on the top-right side. For the second 2x2 block, when the 2-bit LSB is 01 and 11, the pixels allocated to carry are lower-left, upper-right, upper-left, and lower-right in the 2x2 block for four consecutive frames. Will be present in. In both 2x2 pixel blocks, no pixel is processed if the 2-bit LSB is 00. When the 2-bit LSB is 10, the pixel row of the 4x2 block performs switching between the upper row and the lower row every frame. Other embodiments similar to the first embodiment above are provided in the present invention.

Description

3D dither algorithm {3D DITHER ALGORITHM}

1 is a table showing a general two dimensional dither algorithm;

2A-2C are tables showing a typical three dimensional dither algorithm; And

3A and 3B are tables showing a three-dimensional dither algorithm according to a preferred embodiment of the present invention, and FIGS. 4A and 4B are diagrams for better understanding of a table showing a preferred three-dimensional dither algorithm of the present invention shown in FIG. 3A. admit.

The present invention relates to a dither algorithm and, more particularly, to a dynamic dither algorithm for space-time four-time resolution refinement of an image.

A common dither algorithm for 6-bit source drivers in a thin-film-transistor liquid crystal display (TFT LCD) panel provides a dither function with 8 to 6-bit resolution. This dither function assumes only 8 bits and 8 bits have been synthesized, and handles 256 colors mimi. Typically, two algorithms are commonly used, one of which is 2-dimension (2D) static dither, and the other is 3-dimensional (3D) dynamic dither. In general, in both algorithms, blocks each consisting of 2x2 pixels are observed in the image. Two-dimensional dither algorithms provide a constant pattern for carry between observed blocks in the image and achieve low quality, while three-dimensional dither provides more flexibility in carrying patterns Improves defects in dimensional dither. However, three-dimensional dither must be well designed so that the moving or edges of the pixels in the image of the frame are not dithered. Carry and observation blocks will be described in detail below.

To briefly describe the two-dimensional dither algorithm, referring to FIG. 1, 8 to 6 bit dither is illustrated in FIG. 1. However, this is only one example, and 10 to 8 bits, 8 to 6 bits, or 6 to 4 bit dithers are executed in this manner. Basically, the concept of dither is to add one carry (one carry equals 4 for a 2-bit least significant bit) to a valid 6-bit MSB (most significant bit) at the appropriate location and time. First, when the 2-bit LSB is 00, no carry occurs in the 6-bit MSB between the four pixels of the observation block, which includes the active pixel. Next, a carry is added to the 6-bit MSB of any of the 4 pixels when the 2-bit LSB is 01. Then, when the 2-bit LSB is 10, a carry is added to the 6-bit MSB of two of the four pixels. As a result, the average extra pixel value is two. Finally, when the 2-bit LSB is 11, a carry is added to the 6-bit MSB of three of the four pixels, resulting in an average pixel value of three. As described above, the two-dimensional dither algorithm is easy to implement in that the carry can be added only in the spatial domain. An example of a two-dimensional dither method is shown in the table of FIG. 1.

In the two-dimensional dither algorithm, every single pixel with the same value every frame of the same input image is either a dot moire (when LSB is 10) or a quad dot (LSB) on the screen. Observed at a constant pattern, such as 01 or 11). Thus, this static method is rarely used for color dithering.

The conventional three-dimensional dither algorithm, which augments the two-dimensional dither algorithm, is described herein. The three-dimensional dither provides not only dither in the spatial domain but also dither in the temporal domain. Dynamic dithering to the screen, especially in panels with fast response, the movement of pixels is observed. The basic observation unit of the general three-dimensional dither method is also a 2x2 block containing four pixels. When each LSB 2-bit is 10, the general three-dimensional dither method moves the pixels moving left and right in the first horizontal scan line and the second horizontal scan line. ), The movement of pixels moving to the right and left is provided. When each LSB 2-bit is 01 or 11, the general three-dimensional dither method has three basic approaches.

The first conventional method for three-dimensional dither is to perform the movement of the pixel to top-left, top-right, bottom-right, and bottom-left for all 2x2 blocks of consecutive frames. A second conventional method for three-dimensional dither is to perform the movement of the pixel to top-left, bottom-right, bottom-left, and top-right for all 2x2 blocks of consecutive frames. The third conventional method for three-dimensional dither is to perform the movement of the pixel to the top-left, bottom-left, top-right, and bottom-right for all 2x2 blocks of consecutive frames. The tables shown in FIGS. 2A to 2C show methods performed in the general three-dimensional method as described above.

A common problem with 3D dither is that there is no problem when LSB is 10, but when LSB is 01, the horizontal lines move when the first method (Fig. 2a) or the second method (Fig. 2b) is performed. Is the point. And, when all 2x2 blocks move in the same direction, the edge is dithered in the horizontal gray-256 pattern if the first or third (FIG. 2C) scheme is applied. Thus, the problem of three-dimensional dither is found in all three of the above manner.

An object of the present invention is to provide a three-dimensional dither algorithm that can eliminate the movement of horizontal and / or vertical lines between frames on a screen.

Another object of the present invention is to provide a three-dimensional dither algorithm that can prevent the edge of the frame dithering on the screen.

To solve the problem of dynamic dither, the observed unit is extended to a 4x2 block of 8 pixels so that the pixels can be moved more complicated. The period of dynamic dither is invariably composed of four frames. The three-dimensional dither algorithm according to the present invention is as follows.

When the 2-bit LSB is 10, the pixel shift is always up and down. If the average value of 2 is provided on the screen, the problem as described above does not occur. When the 2-bit LSB is 01 or 11, the shift of the pixel (carry for 01 and not carry for 11) is above-left, below-right, below-left, and within the first 2x2 block. Move sequentially to the upper-right side, and sequentially move to the lower-left, upper-right, upper-left, and lower-right within the next 2x2 block to form a 4x2 block. In this type of movement, an average value of 1 or 3 is observed on the screen. Since no horizontal line stays at the same location every two frames, no movement of pixels or lines on the screen is observed on the screen. In addition, since the basic observation unit according to the present invention is a 4x2 block and the surrounding 2x2 blocks perform different operations, the edges are not dithered. Therefore, the three-dimensional dither problem can be solved by the first embodiment of the present invention.

The dither algorithm according to another embodiment of the present invention is similar to the method described above, but the carry element movement rule is opposite to that performed previously. When the 2-bit LSB is 10, the pixel movement always moves up and down as in the first embodiment described above. As a result, an average value of 2 is provided on the screen. When the 2-bit LSB is 01 or 11, the shift of the pixel (which is carry for 01 and not carry for 11) is top-left, top-right, bottom-left, and bottom within the first 2x2 block. Move sequentially to the right side, and sequentially move to the bottom-left, bottom-right, top-left, and top-right side within the next 2x2 block, to constitute a 4x2 block. In this type of movement, an average value of 1 or 3 is observed on the screen. Since no horizontal line stays at the same location every two frames, no movement of pixels or lines is observed on the screen. In addition, since the basic observation unit according to the present invention is a 4x2 block and the surrounding 2x2 blocks perform different operations, the edges are not dithered. Therefore, the three-dimensional dither problem can be solved by the second embodiment of the present invention.

In the above, the problems with the prior art and the advantages of the present invention have been briefly described. It will be apparent to those skilled in the art that various modifications and equivalent other embodiments are possible from the following description and drawings, and from the claims.

(Example)

3A and 3B, tables representing a three-dimensional dither algorithm in accordance with a preferred embodiment of the present invention are presented.

In a preferred embodiment of the present invention, the observation unit is extended to a 4x2 block of 8 pixels so that the pixels can be moved more complicated. The period of dynamic dither is invariably composed of four frames. The three-dimensional dither algorithm according to the present invention is as follows.

Referring first to Figure 3A, when the 2-bit LSB is 10, the pixel movement is always moved up and down. When the 2-bit LSB is 10, the pixel movement is always moved up and down. If an average of 2 is provided on the screen, the problem described above does not occur. When the 2-bit LSB is 01 or 11, the shift of the pixel (carry for 01 and not carry for 11) is above-left, below-right, below-left, and within the first 2x2 block. Move sequentially to the upper-right side, and sequentially move to the lower-left, upper-right, upper-left, and lower-right within the next 2x2 block to form a 4x2 block. In this type of movement, an average value of 1 or 3 is observed on the screen. Since no horizontal line stays at the same location every two frames, no movement of pixels or lines is observed on the screen. In addition, since the basic observation unit according to the present invention is a 4x2 block and the surrounding 2x2 blocks perform different operations, the edges are not dithered. Therefore, the three-dimensional dither problem can be solved by the first embodiment of the present invention.

Referring to FIG. 3B, the dither algorithm according to another embodiment of the present invention is similar to the above-described method, but the carry element movement rule is opposite to that performed previously. When the 2-bit LSB is 10, the pixel movement is always moved up and down as in the first embodiment described above. As a result, an average value of 2 is provided on the screen. When the 2-bit LSB is 01 or 11, the shift of the pixel (which is carry for 01 and not carry for 11) is top-left, top-right, bottom-left, and bottom within the first 2x2 block. Move sequentially to the right side, and sequentially move to the bottom-left, bottom-right, top-left, and top-right within the next 2x2 block to form a 4x2 block. In this type of movement, an average value of 1 or 3 is observed on the screen. Since no horizontal line stays in the same position every two frames, no movement of pixels or lines is observed on the screen. In addition, since the basic observation unit according to the present invention is a 4x2 block and the surrounding 2x2 blocks perform different operations, the edges are not dithered. Therefore, the three-dimensional dither problem can be solved by the second embodiment of the present invention.

4A and 4B are diagrams for better understanding of a table representing a preferred three-dimensional dither algorithm of the present invention shown in FIG. 3A. As described above, the observation unit is extended to a 4x2 block of 8 pixels so that the pixels can be moved more complicated. The 4x2 block includes 2 * 2 first subunits and 2 * 2 second subunits. Each of the first and second sub units consists of an upper-left pixel, an upper-right pixel, a lower-left pixel, and a lower-right pixel. 4A shows first to third operations in a 2 × 2 first subunit. If the 2-bit LSB of the input image pattern is 00, no operation is performed. If the 2-bit LSB of the input image pattern is 01, the carry values of the top-left, top-right, bottom-left, and bottom-right pixels of the first subunit are 1, 0, 0, 0 in the first frame. 0, 0, 0, 1 in the second frame, 0, 0, 1, 0 in the third frame, and 0, 1, 0, 0 in the fourth frame. Here, '1' is a carry and '0' is a non-carry. In other words, if the 2-bit LSB of the input image pattern is 01, the 2-bit LSB of the upper-left pixel, the lower-right pixel, the lower-left pixel and the upper-right pixel in the first sub unit repeatedly every 4-frames. A first operation is performed to add a carry, that is, 1 to the remaining bits except for and to add a non-carry, that is, 0 to the remaining bits except for the 2-bit LSB of the remaining pixels in the first subunit. If the 2-bit LSB of the input image pattern is 10, the carry-ie, 1 is added to the pixels of the lower row and the pixels of the upper row in the first sub-unit repeatedly every two-frames, and 2 of the remaining pixels in the first sub-unit. A second operation is performed that adds non-carry, i.e., 0, to the remaining bits except the -bit LSB. If the 2-bit LSB of the input image pattern is 11, then every 4-frames except for the 2-bit LSB of the top-left pixel, bottom-right pixel, bottom-left pixel and top-right pixel in the first sub-unit repeatedly A third operation is performed to add a non-carry, i.e., zero, to the bits and carry, i.e., one, on the remaining bits except for the 2-bit LSB of the remaining pixels in the first subunit. 4B shows the fourth to sixth operations in the 2 × 2 second subunit. If the 2-bit LSB of the input image pattern is 00, no operation is performed. If the 2-bit LSB of the input image pattern is 01, the carry values of the top-left, top-right, bottom-left and bottom-right pixels of the second sub-unit are 0, 0, 1, 0 in the first frame. , In the second frame, 0, 1 0, 0, in the third frame 1, 0, 0, 0, and in the fourth frame 0, 0, 0, 1. Here, '1' is carry and '0' is non-carrie. In other words, if the 2-bit LSB of the input image pattern is 01, the 2-bit LSB of the lower-left pixel, the lower-right pixel, the upper-left pixel and the upper-right pixel in the second sub unit repeatedly every 4-frames. A fourth operation is performed in which a carry, i.e., is added to the remaining bits except for, and a non-carry, i.e., zero is added to the remaining bits except for the 2-bit LSB of the remaining pixels in the first subunit. If the 2-bit LSB of the input image pattern is 10, every two-frames repeatedly carry, i.e., add 1 to the pixels in the lower row and the pixels in the upper row in the second sub-unit, and 2 of the remaining pixels in the second sub-unit. A fifth operation is performed that adds non-carry, i.e., 0, to the remaining bits except the -bit LSB. If the 2-bit LSB of the input image pattern is 11, then every 4-frames except for the 2-bit LSB of the lower-left pixel, upper-right pixel, upper-left pixel, and lower-right pixel in the first sub-unit repeatedly A sixth operation is performed to add a non-carry, i.e., zero, to the bits and carry, i.e., one, to the remaining bits except for the 2-bit LSB of the remaining pixels in the first subunit. In the above, the problems with the prior art and the advantages of the present invention have been briefly described. It will be apparent to those skilled in the art that various modifications and equivalent other embodiments are possible from the following description and drawings, and from the claims.

As described above, the optimum embodiment has been disclosed in the drawings and the specification. Although specific terms have been used herein, they are used only for the purpose of describing the present invention and are not intended to limit the scope of the invention as defined in the claims or the claims. Therefore, those skilled in the art will understand that various modifications and equivalent other embodiments are possible from this. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

According to the present invention as described above, it is possible to remove the movement of the horizontal and / or vertical lines between the frames on the screen, it is possible to prevent the edges of the frames to be dithered.

Claims (6)

For four-time resolution refinement of an image, comprising a plurality of frames, each of which is scanned into a plurality of pixels of a plurality of bits, and an observation unit consisting of a 4x2 block of pixels, The viewing unit comprises a first subunit and a second subunit, each of the subunits comprising a 2x2 block of pixels, the 2x2 block being an upper-left pixel, an upper-right pixel, a lower-left pixel. And a lower-right pixel, wherein the 2x2 block is an upper row of the upper-left pixel and the upper-right pixel, and a lower row of the lower-left pixel and the lower-right pixel. For a dither algorithm that includes a lower row: In said respective pixels of said first subunit of said viewing unit, If 2-bit LSB (least significant bits) is 00, no action is taken. If the 2-bit LSB is 01, a first operation is performed in which the upper-left pixel, the lower-right pixel, the lower-left pixel, and the upper-right pixel are repeated in a 4-frame period in a continuous form. Doing, If the 2-bit LSB is 10, a second operation is performed in which pixels in the lower row and pixels in the upper row are repeated in a continuous form every 2-frame periods. If the 2-bit LSB is 11, perform the same third operation as that of the first operation; And In the respective pixels of the second sub-unit of the viewing unit, If the 2-bit LSB is 00, no operation is performed. If the 2-bit LSB is 01, a fourth operation of repeating the lower-left pixel, the upper-right pixel, the upper-left pixel, and the lower-right pixel every successive four-frame periods; Then, If the 2-bit LSB is 10, the fifth operation is repeated in the same continuous form as the second operation for each of the 2-frame periods. And if the 2-bit LSB is 11, the dither algorithm performs a sixth operation repeated in the same continuous form as the fourth operation in each of the four-frame periods. The method of claim 1, The repeating pixels in the first operation are carry; The repeated pixels in the second operation are carry; The repeating pixels in the third operation are non-carry; The repeated pixels in the fourth operation are carry; The repeated pixels in the fifth operation are carry; And And wherein the repeated pixels in the sixth operation are non-carried. The method of claim 2, The first and second operations add 1 to the remaining bits except the 2-bit LSB of each of the repeated pixels in the first sub unit, and the remaining pixels except the repeated pixels in the first sub unit. Add 0 to the remaining bits except the 2-bit LSB The third operation adds 0 to the remaining bits except for the 2-bit LSB of each of the repeated pixels in the first sub unit, and adds zero to the remaining bits except for the repeated pixels in the first sub unit. Add 1 to the remaining bits except 2-bit LSB; The fourth and fifth operations add 1 to the remaining bits except the 2-bit LSB of each of the repeated pixels in the second sub unit, and the remaining pixels except the repeated pixels in the second sub unit. Add 0 to the remaining bits except the 2-bit LSB And The sixth operation adds 0 to the remaining bits except for the 2-bit LSB of each of the repeated pixels in the second sub unit, and adds zero to the remaining bits except for the repeated pixels in the second sub unit. A dither algorithm characterized by adding 1 to the remaining bits except for the 2-bit LSB. For four-time resolution refinement of an image, comprising a plurality of frames, each of which is scanned into a plurality of pixels of a plurality of bits, and an observation unit consisting of a 4x2 block of pixels, The viewing unit comprises a first subunit and a second subunit, each of the subunits comprising a 2x2 block of pixels, the 2x2 block being an upper-left pixel, an upper-right pixel, a lower-left pixel. And a lower-right pixel, wherein the 2x2 block is an upper row of the upper-left pixel and the upper-right pixel, and a lower row of the lower-left pixel and the lower-right pixel. For a dither algorithm that includes a lower row: In said respective pixels of said first subunit of said viewing unit, If the 2-bit LSB (least significant bits) is 00, no operation is performed. If the 2-bit LSB is 01, a first operation of repeating the upper-left pixel, the upper-right pixel, the lower-left pixel, and the lower-right pixel in a continuous form every 4-frame periods is performed. Doing, If the 2-bit LSB is 10, the second operation is repeated in a continuous form in the lower row and the upper row every two-frame periods. If the 2-bit LSB is 11, perform the same third operation as that of the first operation; And In the respective pixels of the second sub-unit of the viewing unit, If the 2-bit LSB is 00, no operation is performed. If the 2-bit LSB is 01, a fourth operation in which the lower-left pixel, the lower-right pixel, the upper-left pixel, and the upper-right pixel are repeated in a continuous form every 4-frame periods; Then, If the 2-bit LSB is 10, the fifth operation is the same as the second operation. And if the 2-bit LSB is 11, perform the same sixth operation as that of the fourth operation. The method of claim 4, wherein The repeating pixels in the first operation are carry; The repeated pixels in the second operation are carry; The repeating pixels in the third operation are non-carry; The repeated pixels in the fourth operation are carry; The repeated pixels in the fifth operation are carry; And And wherein the repeated pixels in the sixth operation are non-carried. The method of claim 5, wherein The first and second operations add 1 to the remaining bits except the 2-bit LSB of each of the repeated pixels in the first sub unit, and the remaining pixels except the repeated pixels in the first sub unit. Add 0 to the remaining bits except the 2-bit LSB The third operation adds 0 to the remaining bits except for the 2-bit LSB of each of the repeated pixels in the first sub unit, and adds zero to the remaining bits except for the repeated pixels in the first sub unit. Add 1 to the remaining bits except 2-bit LSB; The fourth and fifth operations add 1 to the remaining bits except the 2-bit LSB of each of the repeated pixels in the second sub unit, and the remaining pixels except the repeated pixels in the second sub unit. Add 0 to the remaining bits except the 2-bit LSB And The sixth operation adds 0 to the remaining bits except for the 2-bit LSB of each of the repeated pixels in the second sub unit, and adds zero to the remaining bits except for the repeated pixels in the second sub unit. A dither algorithm characterized by adding 1 to the remaining bits except for the 2-bit LSB.

Images