TWI396431B - Method of color components compression - Google Patents

Description

Color component compression method

The present invention relates to a method of compression, and more particularly to a method of compressing a color component.

In the design of a general integrated circuit (IC), the size of the data transmitted between the integrated circuit and the memory is considered to be a major factor limiting the performance of the integrated circuit. In order to reduce the time for transferring data and the space for storing data, when the size of the data is relatively large, the data compression program is usually executed before the data is stored in the memory.

In the general video data compression process, the video data is compressed into a standard compression format, such as H.261, H.263, MPEG-1, MPEG-2 and MPEG-4, and then the compressed video data is stored. Go to a compact disc (CD) or a digital video disc (DVD). When the stored video data needs to be read, the compressed data is first decompressed and then stored in the memory. Since the bit rate of the decompressed data is very high, some simple compression/decompression methods capable of processing data in real time can be applied to the data path between the integrated circuit and the memory.

However, when the video material is compressed, the details of some video data are lost during the aforementioned compression process. Therefore, how to compress video data into smaller data sizes without affecting the quality of video data is a major issue to be solved by integrated circuit manufacturers.

In view of the above, the present invention provides a color component compression method in which two compression modes are used to compress video data, thereby reducing the size of the video material while maintaining the quality of the video material.

In order to achieve the above or other objects, the present invention provides a compression method of a color component. This method first accepts a video material encoded in a 4:4:4 sampling format, which includes data for multiple pixels. Then, in a reduced sampling mode, the color components of the plurality of sets of adjacent pixels in the video data are downsampled, and the first difference of the video data before and after the down sampling is calculated. Then, in a truncation mode, the color component of each pixel in the video data is truncated, and the second difference before and after the video data is truncated is calculated. Finally, the first difference and the second difference are compared, and the difference between the sampling mode or the truncation mode is selected to compress the video data into a 4:2:2 sampling format.

In an embodiment of the invention, the video material is decomposed into a plurality of sets of adjacent pixels, and each set of adjacent pixels includes two pixels.

In an embodiment of the invention, the step of downsampling the color components of each set of adjacent pixels in the video data is performed in two pixels of each set of adjacent pixels, and the color component of one of the pixels is retained. And ignore the color component of another pixel.

In an embodiment of the invention, the step of truncating the color component of each set of adjacent pixels in the video material is to cut off the last four bits of the color component of each pixel in the video material.

In an embodiment of the invention, the step of truncating the color components of each set of adjacent pixels in the video data further includes calculating a decimal value of the last four bits of the color component of each pixel, and Determine if this decimal value is greater than 8. If the decimal value is greater than 8, the fifth bit of the color component is incremented by one.

In an embodiment of the invention, after compressing the video data into the 4:2::2 sampling format, the method further stores the compressed video data into the memory.

In an embodiment of the invention, before the compressing the compressed video data into the memory, the method further includes storing an indicator in the last bit of the luminance of each pixel, wherein the indicator record is selected for use. A compression mode that compresses the color components of a pixel.

In an embodiment of the invention, after the compressing the video data into the memory, the method further comprises decompressing the compressed video data in the memory.

In an embodiment of the invention, the step of decompressing the compressed video data in the memory first accesses an index of each pixel of the compressed video data from the memory, and determines a compression mode of the color component of the compressed pixel. . Assuming that the compression mode is determined to reduce the sampling mode, the weighted sum of the color components of the neighboring pixels is calculated (ie, the color component of the left pixel and the color square of the right pixel are summed and divided by two), and the weighted sum is used as per The color component of a group of pixels between adjacent sets of pixels. If the compression mode is determined to be the truncation mode, the recorded color component of each pixel is shifted by four bits and the last four bits of the color component are padded with zeros. Finally, the decompressed video data is output in a 4:4:4 sampling format.

In an embodiment of the invention, before outputting the decompressed video data in a 4:4:4 sampling format, the method further checks whether the two neighboring pixels adjacent to each pixel are decompressed with the same brightness but The brightness of this pixel after decompression is different. If this is the case, the color component of this pixel will be modified to the color component of the neighboring pixel.

In an embodiment of the invention, the color component described above includes a U color component and a V color component.

The present invention directly reduces the sampling and truncation operations for the received video data, and compares the compressed video data with the original video data to determine the best mode for compressing the video data. In this way, the size of the color component of the video material can be reduced and the quality of the video material can be maintained within an acceptable range. Based on the above, the present invention can increase the transmission rate of video data, and can also reduce the space for storing video data.

The above described features and advantages of the present invention will be more apparent from the following description.

Color video data is usually transmitted in the form of luminance and color components, where the brightness retains 90% of the original video data. When the color video data is subjected to distortion compression, the distortion of the brightness is easily recognized by the human eye. Therefore, the present invention provides a method of compressing color components.

Video data is usually transmitted in real 4:4:4 sampling format or in a pseudo 4:4:4 sampling format resulting from 4:2:2 interpolation. Although the two types of video data appear to be in the 4:4:4 sampling format, the characteristics of the two video materials are still different. Therefore, if both video data are considered as interpolated, directly sampled 4:4:4 sampling format data and stored in memory, this will detract the data in true 4:4:4 sampling format. Characteristics in the middle.

FIG. 1 is a schematic diagram of a method of compressing a color component according to an embodiment of the invention. Referring to FIG. 1, this embodiment uses two compression methods to compress video data. One of the methods is to directly perform a down-sampling method, which preserves the U component of 8 bits and the V component of 8 bits in any two pixels, and is suitable for video with less video component changes. The data is compressed. Another method is a truncation method, which only retains the U component of the 4-bit and the V component of the 4-bit of each pixel, and is suitable for compressing video data having more video component changes.

Referring to FIG. 1, during the compression 110 process, when the video data sampled at the 4:4:4 data bit rate is received, the direct down sampling mode 112 and the truncation mode 114 are performed on the video data. . The compressed data is used as the basis for the mode selection 116, which includes calculating the difference between the compressed data and the original data, and selecting to directly reduce the sampling mode 112 and the truncating mode 114 with less difference to compress the video data into 4:2:2 sampling format. Finally, the compressed data sampled at the 4:2:2 data bit rate is stored in the memory 120.

Then, when the read command is received, the compressed video data stored in the memory 120 is retrieved, and the decompressed 130 program is performed on the accessed compressed video data to restore the video data to 4:4:4 data bit rate. Finally, the video data is output in the normal 4:4:4 sampling format. To describe the use of the above architecture to compress and decompress the color components of the video material, two embodiments are provided below to illustrate.

2 is a flow chart of a method for compressing a color component according to an embodiment of the invention. Referring to FIG. 2, this embodiment describes a process of compressing video data, wherein the video data is originally sampled using a data bit rate of 4:4:4.

First, video data encoded in a 4:4:4 sampling format is received (step S202), wherein the video material can be decomposed into a plurality of sets of adjacent pixels and each set of adjacent pixels includes two pixels.

In a reduced sampling mode, the color components of the adjacent pixels in the video data are downsampled, and the first difference of the color components of the video data before and after the sampling is reduced (step S204). More precisely, in the reduced sampling mode, among the two pixels of each set of adjacent pixels, the color component of one of the pixels is retained, and the color component of the other pixel is ignored. When the first difference is calculated, the video data sampled into the 4:2:2 sampling format is reduced to the video data sampled in the 4:4:4 sampling format, and then the calculated 4:4:4 sampling format is calculated. The difference between the color component of the video material and the color component of the received original video material, and the difference is taken as the first difference.

Meanwhile, in a truncation mode, the color component of each pixel in the video material is truncated and the second difference of the color components of the video material before and after being truncated is calculated (step S206). More precisely, in the truncation mode, the last four bits of the color component of each pixel in the video material are truncated. When the second difference is to be calculated, the video data truncated to 4 bits is expanded to 8 bits, and then the color component of the video data of the 4:4:4 sampling format is obtained and the original video received is received. The difference between the color components of the data, and this difference is taken as the second difference.

In this embodiment, it is worth mentioning that when the color component of each pixel in the video data is truncated, the compression method of the color component of the embodiment further calculates the decimal value of the last four bits of the color component of each pixel and Determine if this decimal value is greater than 8. If the decimal value is greater than 8, the fifth bit of the color component is incremented by one. Thus, the truncated color component of each pixel will be closer to the original color component because the fifth bit of the truncated color component will be unconditionally rounded up or unconditionally rounded down.

Finally, comparing the first difference value and the second difference value calculated in the foregoing step to select to reduce the difference between the sampling mode or the truncation mode, and compress the video data into the 4:2:2 sampling format (step S208). More precisely, if the first difference is smaller than the second difference, then the sampling mode is selected to compress the video data into a 4:2:2 sampling format. Conversely, if the second difference is smaller than the first difference, then the truncation mode is selected to compress the video data into a 4:2:2 sampling format.

With the above compression method, the size of the video material can be relatively reduced without losing the details of the color component of the video material. Therefore, the transmission rate of the video data will increase, and the performance of the integrated circuit to access the video data in the memory will also increase.

It is worth mentioning that the compressed video data is also transmitted and stored in the memory. In order to identify the mode in which the video material is actually compressed, this embodiment also stores an indicator in the last bit of the brightness of each pixel, wherein the indicator records the mode that was originally selected to compress the color component of the video material. Accordingly, an embodiment is provided below to illustrate a method of decompressing video material based on the aforementioned criteria for recording.

FIG. 3 is a flow chart of a method for decompressing color components of video data according to an embodiment of the invention. Referring to FIG. 3, this embodiment describes a process of decompressing video data. The video data is downsampled or truncated into a 4:2:2 sampling format, and the original compressed mode is recorded by the indicator, and stored in the memory. in vivo.

First, an index of each pixel in the compressed video material is accessed from the memory (step S302) to determine a mode used when compressing the color component of the pixel. The indicator can be obtained by accessing the last bit of the brightness of each pixel and used to identify the mode that the pixel was originally compressed. For example, if the index read is 0, the compression mode of the video data is recognized as a direct down sampling mode. If the index read is 1, the compression mode of the video data is recognized as the truncation mode.

If the compression mode of the video data is determined to directly reduce the sampling mode, the weighted sum of the color components of the adjacent pixels may be used as the color component of the pixel between each set of adjacent pixels (step S306), thereby restoring at 4 : 4:4 video data in the sample format. More precisely, the color component of the left pixel and the right pixel of each target pixel is added and divided by two, that is, (the color component of the left pixel + the color component of the right pixel)/2, and the weighted sum is used as the target. The color component of the pixel.

If the compression mode of the video data is determined to be the truncation mode, the color component of each pixel recorded is shifted by four bits, and the last four bits are padded with zeros (step S308), thereby restoring 4:4. : 4 video format in sample format.

It is worth mentioning that, in order to reduce the significant difference in the brightness of adjacent pixels that may occur during the decompression process, after the decompression is completed, the embodiment also checks whether the brightness of the two pixels adjacent to each pixel is decompressed. The same but different from the decompressed luminance of the pixels (step S310). If the decompressed luminances of the two adjacent pixels are the same but the decompressed luminances of the pixels are different, the color components of the pixels are modified to the color components of the adjacent pixels (step S312).

More precisely, for one target pixel, the embodiment also checks one pixel on the left and right sides of the target pixel, and determines whether the brightness of the two adjacent pixels is the same. If the brightness of two adjacent pixels is the same, the embodiment further determines whether the brightness of the target pixel is the same as the brightness of the two adjacent pixels. If the brightness of the target pixel is different from the brightness of the two adjacent pixels, it can be determined that there is a significant difference in brightness on the target pixel. Therefore, in this embodiment, the color component of the target pixel is replaced by the color components of two adjacent pixels to make the compressed video data smoother.

Finally, after the "smoothing" process of steps S310 and S312, the decompressed video data is output in a 4:4:4 sampling format (step S314). According to the decompression method described above, the video material of the 4:2:2 sampling format can be restored to the video material of the 4:4:4 sampling format, and the details of the color components of most of the video data can be retained.

In summary, the present invention provides a method for compressing color components by directly reducing the sampling mode and the truncation mode of the received video data to find a mode with less detail loss to compress the video data. Reduce the space for storing video data and still preserve the quality of video data. In addition, only one bit of the brightness of each pixel is used as an index for recording the two modes, and by reading the index, the mode for compressing the color component of the video data can be recognized, and can be used. Properly decompress video data into a 4:4:4 sampling format.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims.

110. . . compression

112. . . Reduce sampling

114. . . Truncated

116. . . Mode selection

120. . . Memory

130. . . unzip

S202~S208. . . Each step of the compression method of the color component according to an embodiment of the present invention

S302~S314. . . Each step of a method for decompressing color components of video data according to an embodiment of the present invention

FIG. 1 is a schematic diagram of a method of compressing a color component according to an embodiment of the invention.

2 is a flow chart of a method for compressing a color component according to an embodiment of the invention.

FIG. 3 is a flow chart of a method for decompressing color components of video data according to an embodiment of the invention.

S202~S208. . . step

Claims (11)

A method for compressing a color component, comprising the steps of: accepting a video data encoded in a 4:4:4 sampling format, the video data comprising data of a plurality of pixels; and in a reduced sampling mode, the amount of the video data The color component of the group of adjacent pixels is subjected to down sampling, and calculating a first difference of the color components of the video data before and after the sampling is reduced; in a truncation mode, the color components of each pixel in the video data are truncated And calculating a second difference of the color components of the video data before and after the truncation; and comparing the first difference with the second difference, and selecting the reduced sampling mode and the truncated mode to have a smaller difference To compress the video data into a 4:2:2 sampling format. A method of compressing color components as recited in claim 1, wherein each set of adjacent pixels comprises two pixels. The method for compressing a color component according to claim 1, wherein the step of downsampling the color components of the neighboring pixels of the video data comprises: retaining the one of the two pixels of each set of adjacent pixels These color components are ignored and the color components of the other pixel are ignored. The method for compressing color components according to claim 1, wherein the step of truncating the color components of each pixel in the video material comprises: truncating the last four of the color components of each pixel in the video data Bit. a method of compressing a color component as described in claim 4 of the patent application, The step of truncating the color components of each pixel in the video data further includes: calculating a decimal value of the last four bits of the color component of each pixel; determining whether the decimal value is greater than 8; If the decimal value is greater than 8, the fifth bit of the color component is incremented by one. The method for compressing a color component according to claim 1, wherein after compressing the video data into the 4:2:2 sampling format, the method further comprises: storing the compressed video data in a memory. The method for compressing a color component according to claim 6 , wherein before storing the compressed video data in the memory, the method further comprises: storing an indicator in a last bit of the brightness of each pixel, The indicator record is selected to compress a compression mode of the color components of the video material. The method for compressing a color component as described in claim 6 , wherein after the step of storing the compressed video material in the memory, the method further comprises: decompressing the compressed video material in the memory. The method for compressing a color component according to claim 8 , wherein the step of decompressing the compressed video material in the memory further comprises: accessing, by the memory, each pixel in the compressed video material. An indicator to determine the compressed mode selected to compress the color components of the pixel; if the compressed mode is determined to be the reduced sampling mode, a weighted sum of the color components of the neighboring pixels is used as Each group of adjacent pixels The color components of the pixels between the pixels; if the compression mode is determined to be the truncation mode, shifting the color components of each pixel by four bits and zeroing the last four bits; and the 4: The 4:4 sampling format outputs the decompressed video material. The method for compressing a color component according to claim 9 , wherein before the step of outputting the decompressed video material in the 4:4:4 sampling format, the method further comprises: checking whether two adjacent pixels of each pixel are used. The decompressed brightness is the same but different from the decompressed brightness of the pixel; and if the decompressed brightness of the neighboring pixels is the same but different from the decompressed brightness of the pixel, the color components of the pixel are modified These color components of the neighboring pixels. A method of compressing color components as recited in claim 1, wherein the color components comprise U color components and V color components.