A method for correcting bright and dark lines in LED display units

Abstract

The present invention relates to a method for correcting bright and dark lines in LED display units. Based on the approximately periodic fluctuations in the integrated brightness values ​​of LED display screen regions, the method applies Gaussian filtering to the integrated values ​​of qualified regions to compensate for abnormal brightness integrals in gap regions, thereby calculating correction coefficients on both sides of the gap. During the bright and dark line correction process, the present invention accurately locates and reduces bright and dark line correction errors. The method is applicable to automated correction systems for visual differences in bright and dark lines on both sides of LED display screens caused by physical splicing gaps, thereby improving screen display quality.

Description

Technical Field

[0001] The invention belongs to the technical field of LED display screen acquisition and correction, and relates to a splicing gap correction method of an LED display unit suitable for automatic correction. Background Art

[0002] The traditional calibration method for large-size LED displays requires that a large number of screens be built and then divided into regions for data collection and calibration, including light and dark line correction. This process is cumbersome and requires a lot of manpower. In addition, if damage occurs during transportation or the cabinets are not placed in the specified order, recalibration or individual inspection of the cabinet placement is required, causing unnecessary trouble. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method for correcting bright and dark lines of an LED display unit suitable for automatic correction, which method can save manpower and facilitate correction.

[0004] In order to solve the above technical problems, the method for correcting bright and dark lines in an LED display unit of the present invention is specifically as follows:

[0005] The LED display unit image is acquired by a camera, and the LED display unit image is processed to obtain the pixel coordinates of the intersection points of each module;

[0006] For any two adjacent modules in a row, when correcting the vertical seam, take the horizontal expansion z×w width on both sides of each vertical seam and the vertical integration area A with the middle (x1~x2)×h height. m1n1 All pixel values ​​are used for brightness integration, 5% ≤ x1 ≤ 10%, 5% ≤ x2 ≤ 10%; for any two adjacent modules in a column, when correcting the horizontal seam, take the vertical extension z × h height on both sides of each horizontal seam and the horizontal center (y1 ~ y2) × w width of the integration area B m2n2 All pixel values ​​are used for brightness integration, 5% ≤ y1 ≤ 10%, 5% ≤ y2 ≤ 10%; 85% ≤ z ≤ 95%; w is the pixel width of each module, and h is the pixel height of each module;

[0007] The integration area A m1n1 Divide the vertical sub-integration area into multiple vertical sub-integration areas with width w / s, and connect the pixel means of all vertical sub-integration areas into a line LW in the order of numbering from left to right m1n1 ; Where s is the number of horizontal light points in the module; the integral area B m2n2 Divide the height h / t into multiple horizontal sub-integration areas, and connect the pixel means of all horizontal sub-integration areas in the order of numbering from top to bottom to form a line LH m2n2 ; Where t is the number of vertical light points in the module; using the filter function To LW m1n1 and LHm2n2 Gaussian filtering is performed to obtain LW m1n1 and LH m2n2 The filtered waveform, where For LW m1n1 The ordinate of the point on the filtered waveform corresponding to the i-th vertical sub-integration area, LH m2n2 The ordinate of the point on the filtered waveform corresponding to the j-th horizontal sub-integration area, is the pixel mean of the i-th vertical sub-integration area, is the pixel mean of the jth horizontal sub-integration area; r is the filter radius, which is equal to 1 to 10 pixel sizes; σ is the standard deviation, σ = 1 to 5;

[0008] Defining LW m1n1 The filtered waveform is a horizontal periodic waveform; find the extreme point A on the standard waveform L closest to the intersection point; let the extreme point A correspond to the ath period of the horizontal periodic waveform, the previous extreme point B of the extreme point A corresponds to the bth period of the horizontal periodic waveform, and the next extreme point C of the extreme point A corresponds to the cth period of the horizontal periodic waveform; take the average of the ordinates of each point in the bth period and the ordinate of the corresponding point in the cth period as the corrected ordinate of the corresponding point in the ath period, and repair the waveform of the ath period; let the corresponding point on the horizontal periodic waveform corresponding to the extreme point A be A k , then the corresponding point A k The correction coefficients of the n columns of pixels on both sides of the vertical seam are obtained by dividing the corrected vertical coordinates of the n corresponding points on both sides by the original vertical coordinates. Similarly, the correction coefficients of the pixels on both sides of the horizontal seam can be obtained; 2≤n≤5;

[0009] For any two adjacent modules in a row, the n repair correction coefficients on both sides of the vertical seam are multiplied with the correction coefficients of the n columns of light points on both sides of the vertical seam to obtain the final correction coefficients of the n columns of light points on both sides of the vertical seam; for any two adjacent modules in a column, the m repair correction coefficients on both sides of the horizontal seam are multiplied with the correction coefficients of the m rows of light points on both sides of the horizontal seam to obtain the final correction coefficients of the m rows of light points on both sides of the horizontal seam. The final correction coefficients are used to correct the pixel values ​​of the light points on both sides of the vertical and horizontal seams to complete the brightness seam repair of the LED display unit.

[0010] The pixel width w and height h of the module are calculated based on the camera pixel size occupied by the LED display unit image, the number of modules in each row and the number of modules in each column.

[0011] The pixel width w and height h occupied by the module can also be calculated based on the pixel coordinates of any four adjacent intersection points.

[0012] When correcting vertical seams, it is preferred to take the integral area A of 0.9×w width extended outward on both sides of each vertical seam and (0.1~0.9)×h height in the middle vertically. m1n1 All pixel values ​​are used for brightness integration.

[0013] When correcting transverse seams, it is preferred to take the integral area B of 0.9×h height extended outward on both sides of each transverse seam and the middle (0.1~0.9)×w width horizontally. m2n2 All pixel values ​​are used for brightness integration.

[0014] The filter radius r is preferably equal to 9 pixel dimensions.

[0015] Preferably, σ=3.

[0016] Beneficial effect: According to the law that the integral value of the brightness of the LED display screen area fluctuates approximately periodically, the present invention calculates the correction coefficients on both sides of the gap by Gaussian filtering the integral value of the qualified area to compensate for the abnormal value of the brightness integral of the gap area. This method is applicable to any screen dot spacing and has a good correction effect whether it is a large-scale acquisition of the entire screen or a fixed-point assembly line acquisition.

[0017] The present invention can accurately locate during the bright and dark line correction process, reduce the bright and dark line correction error, and is suitable for the visual difference of bright and dark lines on both sides of the gap caused by the physical splicing gap of the LED display screen of the automatic correction system, thereby improving the screen display effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a working diagram of the automated acquisition and correction platform.

[0019] In the figure: 1. LED display unit; 2. Automated calibration system platform; 3. Camera.

[0020] Figure 2 This is a schematic diagram of a single module of an LED display.

[0021] Figure 3 This is a schematic diagram of the integral mean area selection.

[0022] Figure 4 This is a trend diagram of the change of the mean value of each sub-integral in the vertical seam integration area before and after filtering (the horizontal axis is the vertical sub-integral area number, and the vertical axis is the pixel mean of the vertical sub-integral area).

[0023] Figure 5 This is a trend diagram of the change of the sub-integral mean before and after filtering in the transverse seam integration area (the horizontal axis is the horizontal sub-integral area number, and the vertical axis is the pixel mean of the transverse sub-integral area).

[0024] Figure 6It is a trend diagram of the change of the pixel mean value of each vertical sub-integral area in the integral area before and after the vertical seam repair.

[0025] Figure 7 It is a trend diagram of the change of the pixel mean value of each transverse sub-integral area in the integral area before and after the transverse seam repair. DETAILED DESCRIPTION

[0026] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions relevant to the present invention, rather than all structures.

[0027] In the description of the present invention, unless otherwise expressly specified or limited, the terms "connected," "connected," and "fixed" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integration; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components or interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0028] In the present invention, unless otherwise expressly specified or limited, a first feature being "above" or "below" a second feature may include the first and second features being in direct contact, or may include the first and second features being in contact not directly but through another feature between them. Furthermore, a first feature being "above," "above," or "above" a second feature may include the first feature being directly above or obliquely above the second feature, or may simply mean that the first feature is higher in level than the second feature. A first feature being "below," "below," or "below" a second feature may include the first feature being directly below or obliquely below the second feature, or may simply mean that the first feature is lower in level than the second feature.

[0029] In the description of this embodiment, terms such as "upper," "lower," "left," and "right" are used to refer to positions or locations based on those shown in the accompanying drawings. These terms are intended solely to facilitate description and simplify operation, and are not intended to indicate or imply that the devices or components referred to must be oriented, constructed, or operated in a specific manner. Therefore, they should not be construed as limitations on the present invention. Furthermore, the terms "first" and "second" are used solely for descriptive purposes and have no special meaning.

[0030] The following is an example of an LED display unit with a dot pitch of 1.5875 mm, 2 modules per row, 3 modules per column, s=96 horizontal light points, and t=72 vertical light points, to describe the technical solution of the present invention in detail.

[0031] The method for correcting bright and dark lines in an LED display unit of the present invention is specifically as follows:

[0032] 1. As Figure 1 As shown, the LED display unit is fixed on the automated calibration system platform, and a camera is used to capture an image of the LED display unit. The LED display unit here can refer to a module or cabinet composed of multiple LED modules, or an LED display screen composed of multiple LED cabinets. The LED display unit image is processed to obtain the pixel coordinates of the intersection points of each module seam. The pixel width w and height h occupied by the module, as well as the pixel size occupied by each light point of the LED display unit, are calculated.

[0033] The pixel width w and height h of the module can be calculated based on the camera pixel size occupied by the LED display unit image, the number of modules in each row, and the number of modules in each column. Assuming that the horizontal pixel size of the LED display unit image is W = 1862 and the vertical pixel size is H = 2094, the pixel width w = W / 2 = 931 and the pixel height h = H / 3 = 698, respectively. The pixel size occupied by each light point is w / s = 9.6979≈10.

[0034] The pixel width w and height h of the module can also be obtained by the following method:

[0035] Randomly select four adjacent intersection points, such as Figure 2 As shown, the pixel coordinates are P1(1022,1217),

[0036] P2 (1953, 1215), P3 (1024, 1914) and P4 (1954, 1913); P1 (1022, 1217) and P3 (1024, 1914) are used as positioning points, and P2 (1953, 1215) and P4 (1954, 1913) are used as auxiliary points; the pixel width w and height h of each module are calculated, where

[0037] 2. If Figure 3 As shown, for any two adjacent modules in a row, when correcting the vertical seam, take the integral area A with a width of 931×0.9 (rounded off) on both sides of each vertical seam and a height of 698×0.1 to 698×0.9 (rounded off) in the middle vertically. m1n1 All pixel values ​​are used for brightness integration; for any two adjacent modules in a column, when correcting the horizontal seam, take the vertical extension of 698×0.9 height (rounded off) on both sides of each horizontal seam and the horizontal width of 931×0.1 to 931×0.9 (rounded off) in the middle. m2n2All pixel values ​​are used for brightness integration.

[0038] 3. Integral area A m1n1 The width of the vertical sub-integration area is 931 / 96≈10, which is divided into several vertical sub-integration areas and numbered 1, 2, ...i... from left to right. The pixel mean is taken for each vertical sub-integration area, and the integration area A is m1n1 The pixel means of all vertical sub-integration areas are connected into lines LW according to their numbers. m1n1 , we can observe the trend of pixel mean value change of each column of light points on both sides of the vertical slit; the integration area B m2n2 The height 698 / 72≈10 is evenly divided into several sub-integration areas, and numbered from top to bottom as 1, 2, ...j..., and the pixel mean of each horizontal sub-integration area is taken, and the integration area B is m2n2 The pixel means of all lateral sub-integration areas are connected into lines LH according to their numbers. m2n2 , we can observe the changing trend of the pixel mean value of each row of light points on both sides of the transverse seam.

[0039] 4. Since the results of w / s and h / t in step 3 are generally not integers, the width of each vertical sub-integration area and the height of each horizontal sub-integration area are not the actual number of camera pixels occupied by the light point; Figure 4 、 Figure 5 As shown, LW m1n1 and LH m2n2 It will change periodically, and due to measurement errors and calculation errors, LW m1n1 and LH m2n2 Each point in the LW will deviate slightly from the standard period, so m1n1 and LH m2n2 Gaussian filtering is performed to obtain LW m1n1 and LH m2n2 The waveform after filtering, the filtering function is: in For LW m1n1 The ordinate of the point on the filtered waveform corresponding to the i-th vertical sub-integration area, LH m2n2 The ordinate of the point on the filtered waveform corresponding to the j-th horizontal sub-integration area, is the pixel mean of the i-th vertical sub-integration area, is the pixel mean of the jth horizontal sub-integration area; r is the filter radius, which is equal to 1 to 10 pixel sizes, preferably equal to 9 pixel sizes; σ is the standard deviation, σ1 to 5, preferably σ=3.

[0040] Defining LW m1n1 The waveform after filtering is a horizontal periodic waveform; LH m2n2The filtered waveform is a longitudinal periodic waveform; each period of the transverse periodic waveform roughly corresponds to the period of the standard waveform; assuming that the intersection point between two adjacent modules in any row and two adjacent modules in the next row is P1(1022,1217) (it can be an upper intersection point or a lower intersection point), find the extreme point A on the transverse periodic waveform that is closest to P1(1022,1217); let the extreme point A correspond to the ath period of the transverse periodic waveform, the previous extreme point B of the extreme point A corresponds to the bth period of the transverse periodic waveform, and the next extreme point C of the extreme point A corresponds to the cth period of the transverse periodic waveform; take the average of the ordinates of each point in the bth period and the ordinates of the points corresponding to the cth period as the corrected ordinates of the points corresponding to the ath period, and repair the waveform of the ath period; for example, Figure 4 As shown, the average of the ordinates of point B16 and point C79 is taken as the corrected ordinate of point A49, ..., the average of the ordinates of point B19 and point C82 is taken as the corrected ordinate of point A52, ..., the average of the ordinates of point B46 and point C109 is taken as the corrected ordinate of point A76, and the waveform after repair of the a-th cycle is obtained; the comparison results of the waveform data before and after the correction of the a-th cycle are shown as follows: Figure 6 As shown; LH m2n2 The filtered waveform is a longitudinal periodic waveform; each period of the longitudinal periodic waveform roughly corresponds to the standard waveform period; assuming that the intersection point between two adjacent modules in any column and two adjacent modules in the previous column is P3(1024,1914), find the extreme point F closest to P3(1024,1914) on the standard waveform period, and set the extreme point F to correspond to the fth period of the longitudinal periodic waveform; similarly, repair the fth period waveform, and the comparison results of the waveform data before and after the fth period correction are as follows: Figure 7 As shown in the figure; the spike part that deviates from the filtered waveform is the original waveform data, and the part that is closer to the standard waveform L is the corrected waveform data; assuming that the extreme point A corresponds to the 64th vertical sub-integral area (i.e., A64), the corrected vertical coordinates of the corresponding point A64 corresponding to the extreme point A and the previous corresponding point A63 and the next two corresponding points A65 and A66 are divided by the vertical coordinates before correction to obtain the repair correction coefficients of the 4 columns of light points on both sides of the vertical seam; similarly, the repair correction coefficients of the 4 rows of light points on both sides of the horizontal seam can be obtained; the vertical seam repair correction coefficient matrix is The correction coefficient matrix for transverse joint repair is: The calculation of the correction coefficients for the repair of horizontal and vertical seams is now completed.

[0041] 5. For any two adjacent modules in a row, multiply the four repair correction coefficients on both sides of the vertical seam with the correction coefficients of the four columns of light points on both sides of the vertical seam to obtain the final correction coefficients of the four columns of light points on both sides of the vertical seam; for any two adjacent modules in a column, multiply the four repair correction coefficients on both sides of the horizontal seam with the correction coefficients of the four rows of light points on both sides of the horizontal seam to obtain the final correction coefficients of the four rows of light points on both sides of the horizontal seam; send the final correction coefficients to the control system to complete the brightness repair of the LED display unit.

[0042] The present invention is not limited to the above embodiment. z is within the range of 85% to 95%, x1 is within the range of 5% to 10%, x2 is within the range of 5% to 10%, y1 is within the range of 5% to 10%, and y2 is within the range of 5% to 10%. When z = 0.9, x1 = 0.1, x2 = 0.9, y1 = 0.1, y2 = 0.9, r is equal to 9 pixels, and σ = 3, the effect is the best. When z is less than 0.9, x1 ≥ 0.1, x2 ≤ 0.9, the data volume may be insufficient and the correction effect may be poor. When z is less than 0.9, y1 ≥ 0.1, y2 ≤ 0.9, the data may introduce errors and even have overlapping parts, which deteriorates the correction effect. When r is less than 9 pixels, the effect waveform is not obvious and the period cannot be accurately located. When σ is less than 3 or greater than 3, the waveform is not standard and the period cannot be accurately located.

Claims

1. A method for correcting bright and dark lines in an LED display unit, characterized in that The method is as follows: The LED display unit image is acquired by a camera, and the LED display unit image is processed to obtain the pixel coordinates of the intersection points of each module; For any two adjacent modules in a row, when correcting the vertical seam, take the horizontal expansion z×w width on both sides of each vertical seam and the vertical integration area A with the middle (x1~x2)×h height. m1n1 All pixel values ​​are used for brightness integration, 5% ≤ x1 ≤ 10%, 5% ≤ x2 ≤ 10%; for any two adjacent modules in a column, when correcting the horizontal seam, take the vertical extension z × h height on both sides of each horizontal seam and the horizontal center (y1 ~ y2) × w width of the integration area B m2n2 All pixel values ​​are used for brightness integration, 5% ≤ y1 ≤ 10%, 5% ≤ y2 ≤ 10%; 85% ≤ z ≤ 95%; w is the pixel width of each module, and h is the pixel height of each module; The integration area A m1n1 Divide the vertical sub-integration area into multiple vertical sub-integration areas with width w / s, and connect the pixel means of all vertical sub-integration areas into a line LW in the order of numbering from left to right m1n1 ; Where s is the number of horizontal light points in the module; the integral area B m2n2 Divide the height h / t into multiple horizontal sub-integration areas, and connect the pixel means of all horizontal sub-integration areas in the order of numbering from top to bottom to form a line LH m2n2 ; Where t is the number of vertical light points in the module; using the filter function To LW m1n1 and LH m2n2 Gaussian filtering is performed to obtain LW m1n1 and LH m2n2 The filtered waveform, where For LW m1n1 The ordinate of the point on the filtered waveform corresponding to the i-th vertical sub-integration area, LH m2n2 The ordinate of the point on the filtered waveform corresponding to the j-th horizontal sub-integration area, is the pixel mean of the i-th vertical sub-integration area, is the pixel mean of the jth horizontal sub-integration area; r is the filter radius, which is equal to 1 to 10 pixel sizes; σ is the standard deviation, σ = 1 to 5; Defining LW m1n1 The filtered waveform is a horizontal periodic waveform; find the extreme point A on the standard waveform L closest to the intersection point; let the extreme point A correspond to the ath period of the horizontal periodic waveform, the previous extreme point B of the extreme point A corresponds to the bth period of the horizontal periodic waveform, and the next extreme point C of the extreme point A corresponds to the cth period of the horizontal periodic waveform; take the average of the ordinates of each point in the bth period and the ordinate of the corresponding point in the cth period as the corrected ordinate of the corresponding point in the ath period, and repair the waveform of the ath period; let the corresponding point on the horizontal periodic waveform corresponding to the extreme point A be A k , then the corresponding point A k The correction coefficients of the n columns of pixels on both sides of the vertical seam are obtained by dividing the corrected vertical coordinates of the n corresponding points on both sides by the original vertical coordinates. Similarly, the correction coefficients of the pixels on both sides of the horizontal seam can be obtained; 2≤n≤5; For any two adjacent modules in a row, the n repair correction coefficients on both sides of the vertical seam are multiplied with the correction coefficients of the n columns of light points on both sides of the vertical seam to obtain the final correction coefficients of the n columns of light points on both sides of the vertical seam; for any two adjacent modules in a column, the m repair correction coefficients on both sides of the horizontal seam are multiplied with the correction coefficients of the m rows of light points on both sides of the horizontal seam to obtain the final correction coefficients of the m rows of light points on both sides of the horizontal seam. The final correction coefficients are used to correct the pixel values ​​of the light points on both sides of the vertical and horizontal seams to complete the brightness seam repair of the LED display unit.

2. The method for correcting bright and dark lines of an LED display unit according to claim 1, characterized in that The pixel width w and height h of the module are calculated based on the camera pixel size occupied by the LED display unit image, the number of modules in each row and the number of modules in each column.

3. The method for correcting bright and dark lines of an LED display unit according to claim 1, characterized in that The pixel width w and height h occupied by the module are calculated based on the pixel coordinates of any four adjacent intersection points.

4. The method for correcting bright and dark lines of an LED display unit according to claim 1, characterized in that When correcting vertical seams, take the integral area A of 0.9×w width extended outward on both sides of each vertical seam and (0.1~0.9)×h height in the middle vertically. m1n1 All pixel values ​​are used for brightness integration.

5. The method for correcting bright and dark lines of an LED display unit according to claim 1, characterized in that When correcting the transverse seam, take the integral area B with a height of 0.9×h extended outward longitudinally on both sides of each transverse seam and a width of (0.1~0.9)×w in the middle horizontally. m2n2 All pixel values ​​are used for brightness integration.

6. The method for correcting bright and dark lines of an LED display unit according to claim 1, characterized in that The filter radius r is equal to 9 pixel sizes.

7. The method for correcting bright and dark lines of an LED display unit according to claim 6, characterized in that σ=3.

Images