A virtual pixel rendering method with dynamic weight allocation

Abstract

A virtual pixel rendering method with dynamic weight distribution relates to the field of electronic display technology and solves the problem of color cast in sharp edge display in existing multiplexing algorithms. The virtual pixel rendering method is implemented based on the RGB triangle arrangement structure; the first row of pixels of the original RGB image is copied upward by one row, the last row of pixels is copied downward by one row, and the last column of pixels is copied rightward by one row; when rendering the grayscale value of any sub-pixel, the grayscale values ​​of the sub-pixel channels corresponding to the six surrounding pixels are read and converted to the CIELAB space at the same time; the central pixel is selected according to the color difference of the read pixel, and a weight is dynamically assigned to each pixel according to the color difference between the surrounding pixels and the central pixel; the grayscale values ​​of the six pixels are weighted and summed to obtain the final output sub-pixel grayscale value. The present invention is particularly suitable for image processing tasks that require high-precision color matching and detail retention, and provides a solid foundation for subsequent analysis.

Description

Technical Field

[0001] The present invention relates to the field of electronic display technology, and in particular to a rendering method for dynamic weight distribution. Background Art

[0002] Pixel multiplexing technology is an important technology for improving the perceived resolution of LED displays. It allows each sub-pixel light point to be shared by several surrounding virtual pixels, similar to the method of smoothing and filtering the original video image data and then downsampling it. It can display images that exceed the physical resolution of the display on the LED display, thereby reducing the pixel pitch, increasing the perceived resolution, and enhancing the display effect.

[0003] For the RGB six-light arrangement, due to the dispersed arrangement of sub-pixels and the unweighted averaging reuse algorithm, during the pixel reuse process, sharp edges contain rich high-frequency information. The unweighted averaging reuse algorithm is also a smoothing filter, which will filter out the rich high-frequency information contained in the edge parts. This leads to insufficient brightness when displaying single-width white lines and single-width black lines. Similarly, when displaying multi-width white lines, the edge sub-pixels have insufficient grayscale due to reuse, resulting in color distortion of sharp edges. The most serious problem is that single-pixel pure black (white background) or pure white (black background) vertical lines will directly appear color-inaccurate. Multi-width white lines will appear purple, yellow, and cyan alternately.

[0004] Taking the display of a single-pixel-width white line as an example, when displaying, three rows of sub-pixels (red, green, and blue) need to be mixed to display a white line; under real pixel display, the grayscale values ​​of the red, green, and blue sub-pixels are all 255, such as Figure 1 As shown, a white line with a grayscale value of 255 can be obtained. However, in six-fold virtual display, since each sub-pixel is shared by six pixels, when a white line with a single pixel width is displayed, the grayscale value of each sub-pixel is calculated as follows: , , , since the grayscale value of each sub-pixel becomes 1 / 3 of the original, such as Figure 2 As shown, this will eventually lead to insufficient display brightness.

[0005] When displaying white lines with multiple line widths, taking the display of 4-line width white lines as an example, under real pixel display, 12 rows of sub-pixels (four rows each of red, green, and blue sub-pixels) are required to achieve the ideal display effect, such as Figure 3 However, in the case of six-fold virtual display, only six rows of sub-pixels are needed (two rows each for red, green, and blue). Figure 4 As shown, taking the calculation of the top three rows of sub-pixels as an example, the calculation results are: , , Obviously, the grayscale value of the edge sub-pixel is much lower than that of the center sub-pixel. The line at the edge that should be white becomes a purple line, causing color distortion at the sharp edge and thus color cast. Summary of the Invention

[0006] In order to solve the problem of color cast on sharp edges in existing multiplexing algorithms, the present invention proposes a virtual pixel rendering method with dynamic weight allocation.

[0007] The technical solutions of the present invention are as follows:

[0008] A virtual pixel rendering method with dynamic weight distribution, wherein the virtual pixel rendering method is implemented by an RGB triangle arrangement structure;

[0009] The RGB triangle arrangement structure is as follows:

[0010] The sub-pixels in the same row are of the same primary color, and the sub-pixels in the same column are cyclically arranged with the three primary colors. Each sub-pixel forms a triangle structure with its two adjacent pixels of different primary colors in the previous column, and also forms a triangle structure with its two adjacent pixels of different primary colors in the next column. Each triangle structure corresponds to one pixel, and each sub-pixel is multiplexed by six pixels.

[0011] The virtual pixel rendering method is specifically as follows:

[0012] S1, copy the first row of pixels of the original RGB image up one row, copy the last row of pixels down one row, and copy the last column of pixels to the right one row;

[0013] S2. When rendering the grayscale value of any sub-pixel, read the grayscale values ​​of the sub-pixel channels corresponding to the six surrounding pixels and convert them to CIELAB space;

[0014] S3, selecting a central pixel based on the color difference of the read pixels, and dynamically assigning a weight to each pixel based on the color difference between the surrounding pixels and the central pixel;

[0015] S4. Perform weighted summation of the grayscale values ​​of the six pixels to obtain a final output sub-pixel grayscale value.

[0016] Preferably, in the RGB triangle arrangement structure, two adjacent sub-pixels in the same row are spaced 2L apart, two adjacent sub-pixels in the same column are spaced 2L apart, and the row spacing between adjacent rows and the column spacing between adjacent columns are both L.

[0017] Preferably, L=0.375 mm.

[0018] Preferably, step S2 specifically includes the following sub-steps:

[0019] S21, setting the grayscale values ​​of the remaining two sub-pixel channels in the same pixel to 0;

[0020] S22. Normalize the RGB values ​​to between 0 and 1, and then perform inverse gamma correction on each channel to obtain linear RGB values:

[0021] ,

[0022] Where C represents the normalized RGB value;

[0023] S23. Use the sRGB standard conversion matrix to convert linear RGB values ​​to XYZ values. Use the following matrix to calculate:

[0024] ,

[0025] in, , , Represents the linear values ​​of R, G, and B sub-pixel channels respectively;

[0026] S24. Using standard illuminant D65 as the reference white point, convert the XYZ values ​​into the CIELAB color space. The conversion formula is as follows:

[0027] ;

[0028] Among them, if , ;otherwise ; , , .

[0029] Preferably, step S3 is specifically as follows:

[0030] Given two specific pixels with coordinates in CIELAB color space, and , calculate the color difference ΔE between them:

[0031] ;

[0032] When ΔE < 2.3, the average value of the Lab* components of the two specific pixels in the CIELAB space is calculated as the central pixel value; the color difference between the remaining four pixels and the central pixel value is calculated respectively. If the color difference of two pixels is more than 10, the average value of the grayscale values ​​of the two high color difference pixels is taken as the final output grayscale value; otherwise, the average value of the color difference between the remaining four pixels and the central pixel value is used. Calculate the weights of surrounding pixels ;The weight of the center pixel ;

[0033] When ΔE≥2.3, calculate the total color difference between the two specific pixels and the remaining four pixels respectively, and select the pixel with the smaller total color difference as the central pixel; calculate the color difference between the central pixel and the remaining five pixels respectively. If the color difference of two pixels is greater than 10, take the average grayscale value of the two high color difference pixels as the final output grayscale value; otherwise, take the average grayscale value of the color difference between the remaining five pixels and the central pixel as the final output grayscale value. Calculate the weights of surrounding pixels ;The weight of the center pixel .

[0034] The present invention also provides a data transmission system, which includes a host computer, an HDMI decoding chip, an HDMI encoding chip, a data processing module and a sending card, wherein the host computer is used to send the HDMI video signal to be displayed to the HDMI decoding chip; the HDMI decoding chip is used to decode the HDMI video signal and send the decoded data to the data processing module; the data processing module includes a virtual pixel rendering unit, which is used to apply the virtual pixel rendering method with dynamic weight distribution as described above to perform rendering calculations on each frame of data; the virtual pixel rendering unit internally includes a FIFO memory and a data register, which is used to dynamically store and read data after each frame of data arrives to form a data matrix, and then calculate the data in the data register according to the virtual pixel rendering method to obtain sub-pixel display data, and send the sub-pixel display data to the HDMI encoding chip to re-encode it into an HDMI video signal and output it to the sending card.

[0035] Preferably, the decoded data includes a horizontal synchronization signal, a vertical synchronization signal, an enable signal and RGB grayscale data.

[0036] Preferably, the data processing module is an FPGA processor.

[0037] The present invention also provides a control system, including a receiving card, a driver IC and the above-mentioned data transmission system. The receiving card is used to receive the effective video signal after being boxed by the sending card in the data transmission system, perform electro-optical conversion and brightness and color correction, and then transmit it to the driver IC to drive the display screen.

[0038] An electronic device comprises the control system described above.

[0039] Compared with the prior art, the present invention has the following specific beneficial effects:

[0040] The rendering method provided by the present invention pre-processes the image, expands the image edge, avoids the problem of missing boundary pixels, reduces conditional judgment branches, reduces the algorithm complexity by more than 20%, and effectively improves the operation speed; and introduces CIELAB color space conversion to more accurately reflect human visual perception. In the dynamic weight allocation process, not only the consistency of color is considered, but also the changes in the local area are taken into account, thereby improving the accuracy and stability of the image processing results. The color difference is quantified in the CIELAB color space, and the color difference between the surrounding pixels and the center pixel is dynamically assigned weights, effectively balancing the impact of different pixels on the final output, and suppressing the blurring effect of smoothing filtering on the edge.

[0041] The method provided by the present invention is particularly suitable for image processing tasks that require high-precision color matching and detail preservation, and provides a solid foundation for subsequent analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of sub-pixel grayscale values ​​when a single-pixel-width white line is displayed in the real pixel display described in the background art;

[0043] Figure 2 A schematic diagram of sub-pixel grayscale values ​​when a single-pixel-width white line is displayed under the sixfold virtual display described in the background art;

[0044] Figure 3 Schematic diagram of sub-pixel grayscale values ​​when displaying multi-line width white lines in the real pixel display described in the background art;

[0045] Figure 4 A schematic diagram of sub-pixel grayscale values ​​when displaying multi-line width white lines under the six-fold virtual display described in the background art;

[0046] Figure 5 Schematic diagram of the specific arrangement structure of sub-pixels in Example 1;

[0047] Figure 6 This is a schematic flow chart of the rendering method of the present invention;

[0048] Figure 7 Schematic diagram showing the comparison before and after the expansion of an image in Example 1;

[0049] Figure 8 This is an example diagram of rendering a red sub-pixel in Example 2. DETAILED DESCRIPTION

[0050] In order to make the technical solution of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the specification of the present invention. It should be noted that the following embodiments are only used to better understand the technical solution of the present invention and should not be understood as limiting the present invention.

[0051] Example 1.

[0052] This embodiment provides a virtual pixel rendering method with dynamic weight allocation, which effectively solves the problems caused by the traditional unweighted average rendering algorithm. The virtual pixel rendering method is implemented based on the RGB triangle arrangement structure.

[0053] The RGB triangle arrangement structure is as follows:

[0054] The sub-pixels in the same row are of the same primary color, and the sub-pixels in the same column are arranged in a cyclic pattern of three primary colors. Each sub-pixel forms a triangle structure with the two adjacent pixels of different primary colors in the previous column, and a triangle structure with the two adjacent pixels of different primary colors in the next column. Each triangle structure corresponds to one pixel, and each sub-pixel is multiplexed by six pixels. The specific arrangement structure of the above sub-pixels is as follows: Figure 5 As shown, a single sub-pixel is shared by the surrounding six pixels to increase the perceived resolution.

[0055] The virtual pixel rendering method is specifically as follows:

[0056] S1, copy the first row of pixels of the original RGB image up one row, copy the last row of pixels down one row, and copy the last column of pixels to the right one row;

[0057] S2. When rendering the grayscale value of any sub-pixel, read the grayscale values ​​of the sub-pixel channels corresponding to the six surrounding pixels and convert them to CIELAB space;

[0058] S3, selecting a central pixel based on the color difference of the read pixels, and dynamically assigning a weight to each pixel based on the color difference between the surrounding pixels and the central pixel;

[0059] S4. Perform weighted summation of the grayscale values ​​of the six pixels to obtain a final output sub-pixel grayscale value.

[0060] Figure 6 The figure shows the block diagram of the algorithm proposed in this paper, which mainly consists of two parts: image preprocessing and dynamic weight allocation. First, the arrangement structure targeted by this method has a single box with a resolution of 960*360 under real pixel display, and can reach a resolution of 1920*1080 under virtual display. However, for the sub-pixels at the edge of the box, when they are rendered, there are no corresponding six pixel data in the original image, so this embodiment expands the image. For an input image of 1920*1080, it needs to be expanded to 1921*1082. As shown in the figure, the resolution of the single box is 960*360 under real pixel display, and can reach 1920*1080 under virtual display. However, for the sub-pixels at the edge of the box, when they are rendered, there is no corresponding six pixel data in the original image, so this embodiment expands the image. For an input image of 1920*1080, it needs to be expanded to 1921*1082. Figure 7In the schematic diagram, the blue pixels are the expanded parts, which makes the subsequent calculations more concise and improves the operation speed. In order to better meet the visual effects of the human eye during processing, this embodiment converts the grayscale values ​​of the corresponding sub-pixel channels into CIELAB space. Secondly, the center pixel is selected based on the color difference of the read pixel, and a weight is dynamically assigned to each pixel based on the difference between the surrounding pixels and the center pixel. Finally, the rendering submodule assigns these values ​​to each sub-pixel, and finally outputs the rendered image in RGB form.

[0061] Example 2.

[0062] On the basis of Example 1, in the rendering method of this embodiment, step S2 specifically includes the following sub-steps:

[0063] S21, setting the grayscale values ​​of the remaining two sub-pixel channels in the same pixel to 0;

[0064] Since red, green, and blue are rendered in the same way, we only use the red sub-pixel as an example. Figure 8 As shown in the figure, when rendering the red sub-pixel, it is necessary to read the grayscale values ​​of the red channel corresponding to the six pixels a1, a2, b1, b2, c1, and c2, which are denoted as R1, R2, R3, R4, R5, and R6. Then, converting to the CIELAB space, it is necessary to expand the grayscale values ​​of the six sub-pixels first, that is, set the grayscale values ​​of their blue and green channels to 0. They are denoted as (0, R1, 0), (0, R2, 0), (0, R3, 0), (0, R4, 0), (0, R5, 0), and (0, R6, 0).

[0065] S22. Normalize the RGB values ​​to between 0 and 1, and then perform inverse gamma correction on each channel to obtain linear RGB values:

[0066] ,

[0067] Where C represents the normalized RGB value;

[0068] S23. Use the sRGB standard conversion matrix to convert linear RGB values ​​to XYZ values. Use the following matrix to calculate:

[0069] ,

[0070] in, , , Represents the linear values ​​of R, G, and B sub-pixel channels respectively;

[0071] S24. Convert the RGB color space to the CIELAB color space to more accurately reflect human visual perception. When converting the color space, select D65 as the reference white point and convert the XYZ values ​​to the CIELAB color space. The conversion formula is as follows:

[0072] ;

[0073] Among them, if , ;otherwise ; , , .

[0074] Example 3.

[0075] Based on Example 2, this example specifically defines step S3 as follows:

[0076] After the color space conversion, the color difference ΔE between two pixels is calculated using the directional color difference calculator.

[0077] Given two specific pixels (called b1 and b2) whose coordinates in CIELAB color space are and , these two pixels represent the color features in the local area, and the color difference ΔE between them is calculated:

[0078] ;

[0079] The color difference between b1 and b2 determines how to choose the reference point:

[0080] When ΔE < 2.3, the average value of the Lab* components of the two specific pixels in the CIELAB space is calculated as the central pixel value, that is:

[0081] ,

[0082] ,

[0083] ;

[0084] Calculate the color difference between the remaining four pixels and the central pixel value respectively. If the color difference of two pixels is greater than 10, take the average grayscale value of these two high color difference pixels as the final output grayscale value; otherwise, calculate the weight of the surrounding pixels according to the average color difference between the remaining four pixels and the central pixel value. ,in ;The weight of the center pixel ;

[0085] When ΔE≥2.3, calculate the total color difference between these two specific pixels and the remaining four pixels (i.e., a1, a2, c1, c2) respectively, and select the pixel with the smaller total color difference as the center pixel. That is, if the total color difference between b1 and the surrounding four pixels is smaller than the total color difference between b2 and the surrounding four pixels, then select b1 as the reference point; otherwise, select b2 as the reference point. This process ensures that even in the case of large color differences, a color reference point that is more suitable as the center can be found, thereby reducing errors caused by color mutations. Calculate the color difference between the center pixel and the remaining five pixels respectively. If there are two pixels whose color difference exceeds 10, take the average grayscale value of these two high color difference pixels as the final output grayscale value; otherwise, take the average color difference between the remaining five pixels and the center pixel. Calculate the weights of surrounding pixels ,in, Here we assume that b1 is the center pixel and b2 is the surrounding pixels; the weight of the center pixel is .

[0086] Finally, based on the weights calculated above, combined with the grayscale values ​​of all six pixels, the final output grayscale value is calculated using the weighted average method. Specifically, the grayscale values ​​of all pixels are weighted and summed to obtain the final output sub-pixel grayscale value, as shown in the following formula:

[0087] .

[0088] Example 4.

[0089] This embodiment provides a data transmission system, which applies a rendering method in any of the above embodiments. The data transmission system includes a host computer, an HDMI decoding chip, an HDMI encoding chip, a data processing module and a sending card. The host computer is used to send the HDMI video signal to be displayed to the HDMI decoding chip; the HDMI decoding chip is used to decode the HDMI video signal and send the decoded data to the data processing module; the data processing module includes a virtual pixel rendering unit, which is used to perform rendering calculations on each frame of data using the virtual pixel rendering method with dynamic weight allocation described in any one of Embodiments 1 to 3; the virtual pixel rendering unit internally includes a FIFO memory and a data register, which is used to dynamically store and read data after each frame of data arrives to form a data matrix, and then calculate the data in the data register according to the virtual pixel rendering method to obtain sub-pixel display data, and send the sub-pixel display data to the HDMI encoding chip for re-encoding into an HDMI video signal and output to the sending card.

[0090] The data transmission system for the pixel multiplication display screen provided in this embodiment applies the method described in Example 1, which not only takes into account the consistency of color, but also takes into account the changes in the local area, thereby improving the accuracy and stability of the image processing results. This strategy is particularly suitable for image processing tasks that require high-precision color matching and detail retention, and provides a solid foundation for subsequent analysis. By dynamically adjusting the weights, the impact of different pixels on the final output is effectively balanced, making the processing results more in line with actual needs. It also achieves virtual processing of data on the transmission path between the host computer and the sending card, saving the transmission bandwidth between the sending card and the receiving card, the sending card load, and the data processing capacity of the receiving card, reducing the system's requirements for the sending card and the receiving card, thereby significantly reducing costs; it also reduces the number of wiring between the sending card and the receiving card in the corresponding display screen control device, making the wiring inside the device box regular, reducing the risk of short circuits, and making the device easy to maintain.

[0091] Example 5.

[0092] On the basis of Example 4, in the data transmission system provided by this embodiment, the decoded data includes a horizontal synchronization signal, a vertical synchronization signal, an enable signal and RGB grayscale data.

[0093] Example 6.

[0094] On the basis of Example 4, in the data transmission system provided by this embodiment, the data processing module is an FPGA processor.

[0095] Example 7.

[0096] This embodiment provides a control system, including a receiving card, a driver IC, and the data transmission system as described in the above embodiment. The receiving card is used to receive the valid video signal after being boxed by the sending card in the data transmission system, perform electro-optical conversion and brightness and color correction, and then transmit it to the driver IC to drive the display screen.

[0097] Example 8.

[0098] This embodiment provides an electronic device, which includes the display screen control system described in Example 7.

Claims

1. A virtual pixel rendering method with dynamic weight allocation, characterized in that: The virtual pixel rendering method is implemented by RGB triangle arrangement structure; The RGB triangle arrangement structure is as follows: The sub-pixels in the same row are of the same primary color, and the sub-pixels in the same column are cyclically arranged with the three primary colors. Each sub-pixel forms a triangle structure with its two adjacent pixels of different primary colors in the previous column, and also forms a triangle structure with its two adjacent pixels of different primary colors in the next column. Each triangle structure corresponds to one pixel, and each sub-pixel is multiplexed by six pixels. In the RGB triangle arrangement structure, the interval between two adjacent sub-pixels in the same row is 2L, the interval between two adjacent sub-pixels in the same column is 2L, and the row spacing between adjacent rows and the column spacing between adjacent columns are both L. The virtual pixel rendering method is specifically as follows: S1, copy the first row of pixels of the original RGB image up one row, copy the last row of pixels down one row, and copy the last column of pixels to the right one row; S2. When rendering the grayscale value of any sub-pixel, read the grayscale values ​​of the sub-pixel channels corresponding to the six surrounding pixels and convert them to CIELAB space; S3, selecting a central pixel based on the color difference of the read pixels, and dynamically assigning a weight to each pixel based on the color difference between the surrounding pixels and the central pixel; S4. Perform weighted summation of the grayscale values ​​of the six pixels to obtain a final output sub-pixel grayscale value.

2. The virtual pixel rendering method with dynamic weight allocation according to claim 1, characterized in that: L=0.375mm.

3. The virtual pixel rendering method with dynamic weight allocation according to claim 1 or 2, characterized in that: Step S2 specifically includes the following sub-steps: S21, setting the grayscale values ​​of the remaining two sub-pixel channels in the same pixel to 0; S22. Normalize the RGB values ​​to between 0 and 1, and then perform inverse gamma correction on each channel to obtain linear RGB values: ， Where C represents the normalized RGB value; S23. Use the sRGB standard conversion matrix to convert linear RGB values ​​to XYZ values. Use the following matrix to calculate: ， in, , , Represents the linear values ​​of R, G, and B sub-pixel channels respectively; S24. Using standard illuminant D65 as the reference white point, convert the XYZ values ​​into the CIELAB color space. The conversion formula is as follows: ； Among them, if , ;otherwise ; , , .

4. The virtual pixel rendering method with dynamic weight allocation according to claim 3, characterized in that: Step S3 is as follows: Given the coordinates of two specific pixels in the middle row of the six pixels in the CIELAB color space are and , calculate the color difference ΔE between them: ； When ΔE < 2.3, the average of the L*, a*, and b* components of the two specific pixels in the CIELAB space is calculated as the central pixel value; Calculate the color difference between the remaining four pixels and the central pixel value respectively. If the color difference of two pixels exceeds 10, take the average grayscale value of these two high color difference pixels as the final output grayscale value; Otherwise, the average value of the color difference between the remaining four pixels and the central pixel is used Calculate the weights of surrounding pixels ;The weight of the center pixel ; When ΔE ≥ 2.3, the total color difference between the two specific pixels and the remaining four pixels is calculated respectively, and the pixel with the smaller total color difference is selected as the central pixel. The color difference between the central pixel and the remaining five pixels is calculated respectively. If the color difference of two pixels exceeds 10, the average grayscale value of the two high color difference pixels is taken as the final output grayscale value. Otherwise, the average color difference between the remaining five pixels and the central pixel is calculated. Calculate the weights of surrounding pixels ;The weight of the center pixel .

5. A data transmission system, characterized in that: The data transmission system includes a host computer, an HDMI decoding chip, an HDMI encoding chip, a data processing module and a sending card, wherein the host computer is used to send the HDMI video signal to be displayed to the HDMI decoding chip; the HDMI decoding chip is used to decode the HDMI video signal and send the decoded data to the data processing module; the data processing module includes a virtual pixel rendering unit, which is used to apply the virtual pixel rendering method with dynamic weight allocation according to any one of claims 1 to 4 to perform rendering calculations on each frame of data; the virtual pixel rendering unit internally includes a FIFO memory and a data register, which is used to dynamically store and read out the data after each frame of data arrives to form a data matrix, and then calculate the data in the data register according to the virtual pixel rendering method to obtain sub-pixel display data, and send the sub-pixel display data to the HDMI encoding chip to re-encode it into an HDMI video signal and output it to the sending card.

6. The data transmission system according to claim 5, characterized in that The decoded data includes a horizontal synchronization signal, a vertical synchronization signal, an enable signal and RGB grayscale data.

7. The data transmission system according to claim 5, characterized in that The data processing module is an FPGA processor.

8. A control system, characterized in that: The device comprises a receiving card, a driver IC and a data transmission system as described in any one of claims 5 to 7, wherein the receiving card is used to receive the effective video signal after being binned by the sending card in the data transmission system, perform electro-optical conversion and brightness and color correction, and then transmit it to the driver IC to drive the display screen.

9. An electronic device, characterized in that: The electronic device includes the control system according to claim 8.

Images