Sub-pixel positioning method and device based on edge contour information

Abstract

The invention provides a sub-pixel positioning method and device based on edge contour information. The method comprises the following steps: step 1, modeling a display panel contour; specifically, edge pixel coordinates of the display panel are obtained through a local maximum value; contour pixel coordinates are obtained through convex hull detection, a four-side contour point set is determinedthrough a Hough transform detection straight line algorithm, a quadratic polynomial is adopted to approach the contour of a light-emitting area of the display panel, and corresponding vertexes and side lengths are further solved through four equations established by the four-side contour; 2, meshing the boundary area solved in the step 1 according to the display resolution, establishing a grid index information table, and sequencing by searching a four-side contour point set in the neighborhood of grid index points; and 3, establishing an n-order quadratic homogeneous polynomial model by utilizing the point set indexes and coordinates sorted in the step 2 to realize coordinate calculation of all pixel points. Compared with the traditional method, the method disclosed by the invention has the advantage that sub-pixel-level accurate positioning can be realized only by lightening part of contour pixel points on a screen.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to a sub-pixel positioning method and device based on edge contour information. Background technique [0002] The prior art generally adopts the corner point positioning method to carry out the pixel point positioning of the OLED or μLED screen, including the Harris corner point detection algorithm, the FAST algorithm and the BRJEF algorithm using the checkerboard to find points; The first-order derivative matrix detection method of the degree mainly uses the local self-similarity / autocorrelation, that is, the similarity between the image block in a certain local window and the image block in the window after a small movement in all directions is used for detection. The method of judging whether a candidate point is a corner point based on the FAST algorithm of the accelerated segmentation test is: on a discretized Bresenllam circle with a pixel as the center and a radius...

AI Technical Summary

Problems solved by technology

[0003] However, the pictures targeted by the above methods must be images with continuous corners and no interference of overly bright or overly dark block pixels, while OLED or μLED screens may have bright and dark band patterns, some pixels are not bright, OLED or μLED screens may After the μLED pixels are imaged by the camera, they become a large area of ​​light spots and other problems. Therefore, the existing methods cannot meet the demand for high-precision, stable and robust pixel-level positioning of OLEDs or μLEDs, and in industrial applications. Pixel positioning of this type of screen needs to meet the processing time requirements of assembly line production, and it is expected that the positioning time will be as short as possible

Images