Search

CN-121544621-B - Uniformity statistical method and device for display image and storage medium

CN121544621BCN 121544621 BCN121544621 BCN 121544621BCN-121544621-B

Abstract

The application discloses a uniformity statistical method, a device and a storage medium of a display image, which are used for improving the accuracy of gray uniformity analysis of the shot image of the display. Generating a minimum external rectangular frame in a row/column direction for a normalized image, performing intersection processing and connected domain separation processing according to the minimum external rectangular frame in the row/column direction, calculating gray level statistic information values of a separation region, generating a blank image, replacing gray level values of the blank image with the gray level statistic information values, generating a row curved surface image and a column curved surface image, cycling an array of the blank image, generating a dynamic row-column coordinate array of a neighborhood, generating a vector field image, performing algorithm remapping processing on the normalized image and the vector field image, generating a result image, merging the result image, filling the maximum pixel values of a multichannel image into an image to be processed, generating a target filling image, and performing gray level distribution uniformity analysis according to gray level value data.

Inventors

  • SHI DONGDONG
  • YANG SHUO
  • CUI QIAOQIAO

Assignees

  • 深圳精智达技术股份有限公司

Dates

Publication Date
20260508
Application Date
20260120

Claims (10)

  1. 1. A method for uniformity statistics of a display image, comprising: acquiring a normalized image of an image shot by a display, and generating a minimum circumscribed rectangular frame in a row direction and a minimum circumscribed rectangular frame in a column direction for the normalized image; Performing intersection processing and connected domain separation processing according to the minimum circumscribed rectangular frame in the row direction and the minimum circumscribed rectangular frame in the column direction, and calculating a gray level statistical information value of a separation region; generating a blank image, and replacing the gray value of the blank image with a gray statistical information value; generating a row curved surface image and a column curved surface image according to the replaced blank image; circularly replacing the array of the blank image to generate a dynamic row-column coordinate array of a neighborhood, and generating a vector field image according to the row curved surface image, the column curved surface image and the dynamic row-column coordinate array; carrying out algorithm remapping treatment on the normalized image and the vector field image, and carrying out absolute value difference operation on the remapped image and the normalized image to generate at least 8 result images; Selecting a plurality of result images for each position area respectively, merging the result images into a multi-channel image, filling the maximum pixel value of the multi-channel image into the corresponding area on the image to be processed, and generating a plurality of filled images to be processed, wherein the images to be processed are real number type images with gray values of 0, which are generated in advance; And carrying out fusion and superposition on the filled image to be processed to generate a target filled image, extracting gray value data on the target filled image, and carrying out gray uniformity analysis according to the gray value data.
  2. 2. The uniformity statistical method of claim 1, wherein the step of generating a dynamic row-column coordinate array of a neighborhood from the array of cyclically replaced blank images, generating a vector field image from the row curved image, the column curved image, and the dynamic row-column coordinate array comprises: the origin of the current pixel coordinate is used as the center, the array of the blank image after the circular replacement is used for generating a dynamic row-column coordinate array of the neighborhood, and the dynamic row-column coordinate arrays are respectively recorded; removing intermediate value elements in the dynamic row-column coordinate arrays, and sequentially extracting the rest row-column coordinate arrays according to the index subscript order; Sequentially shifting the row curved surface image and the column curved surface image to adjust the image shift gray value; Performing domain-defining limiting clamping control processing on the displaced line curved surface image and column curved surface image; And performing logical AND operation on the area after the card control, and converting the two real value images after the logical AND operation into vector field images to generate the vector field images.
  3. 3. The uniformity statistical method according to claim 1, wherein the step of selecting a plurality of result images for each location area to be combined into a multi-channel image, filling the maximum pixel value of the multi-channel image into the corresponding area on the image to be processed, and generating a plurality of filled image to be processed comprises: determining a plurality of corner points, determining a corner point area for each corner point, and determining a plurality of related result images for each corner point area; combining a plurality of result images related to the corner areas into a multi-channel image, and solving the maximum value of the pixel values of the corner areas in the multi-channel image; Filling the maximum pixel value corresponding to each corner region into the corresponding region on the new image to be processed, and generating a plurality of filled images to be processed; Determining a plurality of rectangular row and column areas, and respectively determining a plurality of related result images for each rectangular row and column area, wherein the rectangular row and column area comprises a rectangular head row area, a rectangular tail row area, a rectangular head column area and a rectangular tail column area; Combining a plurality of result images related to the rectangular line and column areas into a multi-channel image, and solving the maximum value of pixel values in the multi-channel image at the same coordinate position as the rectangular line and column areas; Filling the maximum pixel value corresponding to each rectangular row and column region into the corresponding region on the new image to be processed, and generating a plurality of filled images to be processed; determining a central area, and acquiring a plurality of result images corresponding to the central area; combining a plurality of result images related to the central area into a multi-channel image, and solving the maximum value of pixel values in the multi-channel image at the same coordinate position as the central area; And filling the maximum pixel value corresponding to the central region into the corresponding region on the new image to be processed, and generating a filled image to be processed.
  4. 4. The uniformity statistic method of claim 1, wherein said step of obtaining a normalized image of a display captured image, generating a row-wise minimum bounding rectangular box and a column-wise minimum bounding rectangular box for said normalized image comprises: Normalizing the acquired display shooting image to generate a normalized image; Calculating the total image segmentation number of the normalized image according to the width and height dimensions of the image shot by the display and a preset segmentation step length; generating a fixed-length array and a fixed-value array according to the total image segmentation number, wherein the array numbers of the fixed-length array range and the fixed-value array are determined by the total image segmentation number; Generating 4 dynamic arrays according to the fixed-length array, the fixed-value array, a preset segmentation step length and the total segmentation number; And generating a minimum circumscribed rectangular frame in the row direction and a minimum circumscribed rectangular frame in the column direction for the normalized image according to the coordinate sequence by using the dynamic array.
  5. 5. The uniformity statistical method of claim 1, wherein the step of generating a row curved image and a column curved image from the replaced blank image comprises: And generating a row curved surface image and a column curved surface image according to the gray scale of the center point of the blank image after replacement and the gray scale change rate in the row-column direction.
  6. 6. The uniformity statistical method of any one of claims 1 to 5, wherein prior to the step of obtaining a normalized image of a display captured image, generating a row-wise minimum bounding rectangle box and a column-wise minimum bounding rectangle box for the normalized image, the uniformity statistical method further comprises: After a carrier camera environment is set in a darkroom, driving and lighting a display through PG lighting equipment, photographing and drawing the display, and generating a display photographing image; performing straight line fitting on the shot image of the display to generate a plurality of fitting straight line equations; calculating the space distance between adjacent straight line intersection points by using the fitting straight line equations; performing image analysis according to the spatial distance and the resolution of the image shot by the display; And when the condition that the image shot by the display does not reach the clamping threshold value is met, carrying out position adjustment on the carrier camera, and re-shooting the image shot by the display until the condition that the image shot by the display reaches the clamping threshold value is met.
  7. 7. The uniformity statistical method of claim 6, wherein the step of generating a plurality of fit straight-line equations by straight-line fitting the display captured image comprises: Determining transition edge points of foreground images and background images of the shot images of the display in all directions; And fitting the transition edge points to generate a plurality of fitting linear equations.
  8. 8. The uniformity statistical method of claim 6, wherein after the step of adjusting the position of the stage camera and re-capturing the display captured image until the display captured image reaches the condition of the snap threshold, and before the step of obtaining the normalized image of the display captured image, generating a row-direction minimum bounding rectangle frame and a column-direction minimum bounding rectangle frame for the normalized image, the uniformity statistical method further comprises: and according to the intersection points of the adjacent straight lines and the preset ideal point pairs, carrying out integral correction transformation on the display shooting images reaching standards through a point pair one-to-one mapping matrix.
  9. 9. A display image uniformity statistics apparatus, comprising: The device comprises an acquisition unit, a display unit and a display unit, wherein the acquisition unit is used for acquiring a normalized image of an image shot by the display unit and generating a minimum circumscribed rectangular frame in a row direction and a minimum circumscribed rectangular frame in a column direction for the normalized image; The first calculation unit is used for carrying out intersection processing and connected domain separation processing according to the minimum circumscribed rectangular frame in the row direction and the minimum circumscribed rectangular frame in the column direction, and calculating gray statistical information values of the separation areas; the first generation unit is used for generating a blank image and replacing the gray value of the blank image with a gray statistical information value; the second generation unit is used for generating a row curved surface image and a column curved surface image according to the replaced blank image; The third generation unit is used for circularly replacing the array of the blank image to generate a dynamic row-column coordinate array of the neighborhood, and generating a vector field image according to the curved surface image, the column curved surface image and the dynamic row-column coordinate array; The fourth generation unit is used for carrying out algorithm remapping processing on the normalized image and the vector field image, carrying out absolute value difference operation on the remapped image and the normalized image, and generating at least 8 result images; A fifth generating unit, configured to select a plurality of result images for each position area, and combine the result images into a multi-channel image, fill the maximum pixel value of the multi-channel image into a corresponding area on the image to be processed, and generate a plurality of filled images to be processed, where the image to be processed is a real number type image with a gray value of 0; The first analysis unit is used for carrying out fusion superposition on the filled image to be processed to generate a target filled image, extracting gray value data on the target filled image and carrying out gray uniformity analysis according to the gray value data.
  10. 10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a program which, when executed on a computer, performs the uniformity statistical method according to any one of claims 1 to 8.

Description

Uniformity statistical method and device for display image and storage medium Technical Field The present application relates to the field of display screen detection, and in particular, to a method and apparatus for uniformity statistics of a display image, and a storage medium. Background With technological innovation, new display technologies (such as MicroLED, flexible folding screens, etc.) are continuously updated iteratively. As a key link of the display screen industry chain, the quality detection technology of the display screen is continuously paid attention to, and aiming at each detection of the surface of the display screen, the quality detection technology of the display screen is a key element for determining the quality of the display screen product. Wherein, optical detection and De-Mura compensation for a display screen are still indispensable links in the panel process. The De-Mura compensation process is based on the gray scale values of each Pattern frame captured by the high resolution camera. Before the algorithm processing is started, the program can be subjected to a series of image preprocessing, so that the gray level difference value of each picture can more truly restore the brightness state and Mura form of the display screen. With the application of more and more fields to display screens, new display screens are continuously rising, and defect detection, uniformity and repeatability testing for the new display screens are also important. Nowadays, in order to adapt to a specific technical field, a specific area is designed to adapt to the actual use of the technical field, so that the projects of defect detection required to be performed on the novel display screen are more and more, and the complexity of defect detection is higher and higher. Especially for micro-display, the product size of the existing micro-display product is smaller, the structure is more precise, and the design mode of the pixel point is more complex. Therefore, when defect detection is performed on the display, uniformity of images acquired by the display needs to reach a preset standard, gray level uniformity analysis is usually required to be performed on images shot by the display, particularly, the display with a folding function and a splicing function is further added, and gray level uniformity analysis on images shot by the display needs to be more accurate, but in reality, gray level uniformity analysis on images shot by the display is more difficult compared with a conventional display screen. Specifically, when an image is shot by a more complex display, the more complex display is more easily interfered by external environment than a conventional display screen, and in order to ensure uniformity of the shot image, a generally selected statistical method is whole-area block absolute value statistics, but such statistical method cannot perform correlation analysis on neighborhood of the shot image of the display, so that accuracy of gray uniformity analysis of the shot image of the display is reduced. Disclosure of Invention The application discloses a uniformity statistical method, a device and a storage medium of a display image, which are used for improving the accuracy of gray uniformity analysis of the shot image of the display. The embodiment of the application provides a uniformity statistical method of a display image, which comprises the steps of obtaining a normalized image of the display image, generating a minimum circumscribed rectangular frame in a row direction and a minimum circumscribed rectangular frame in a column direction for the normalized image, carrying out intersection processing and connected domain separation processing according to the minimum circumscribed rectangular frame in the row direction and the minimum circumscribed rectangular frame in the column direction, calculating gray level statistical information values of separation areas, generating a blank image, replacing the gray level values of the blank image with the gray level statistical information values, generating a row curved surface image and a column curved surface image according to the replaced blank image, circularly replacing an array of the blank image, generating a dynamic row and column coordinate array of a neighborhood, generating a vector field image according to the curved surface image, the column curved surface image and the dynamic row and column coordinate array, carrying out algorithm remapping processing on the normalized image and the vector field image, carrying out absolute value difference operation on the remapped image and the normalized image, generating at least 8 result images, respectively selecting a plurality of result images for each position area to be combined into a plurality of multi-channel images, filling the maximum pixel values of the multi-channel images into corresponding areas on the images to be processed images, generating a to-level data of the to-be-proce