High-precision sub-pixel extraction method for centers of light bars of structured light

Abstract

The invention discloses a high-precision sub-pixel extraction method for the centers of light bars of structured light, and belongs to the technical field of computer vision. The method is characterized in that that the method employs a skeleton refining method to obtain the initial positions of the centers of stripes; solving the normal direction of each position on a skeleton through a mean square gray scale gradient; employing an adaptive threshold value to extract the normal width value of each column of stripes; carrying out Fourier fitting through gray scale data in the stripe normal width value; solving a peak value of a fitting curve, and obtaining the sub-pixel coordinates of the centers of the light bars, and connecting the centers of the light bars to form the center of the light bar. The extracted center of the light bar is higher in precision, and the method also can accurately extract the center of the light boars under the condition that the gray scale of the light bar and the width of the light bar are not uniform.

Description

technical field [0001] The invention relates to an image processing technology, in particular to a machine vision technology, in particular to a high-precision sub-pixel extraction method for the center of a structured light stripe. Background technique [0002] At present, structured light active vision is mainly based on the principle of optical triangulation, and the vision system is composed of a laser generator and a camera as a structured light image sensor. Due to the irregular shape of the surface of the object and the inconsistent reflection characteristics, the image captured by the image sensor containing the structured light bar has an uneven grayscale and width of the light bar. These bring great difficulties to the stripe center extraction of the light stripe image, so the key step in the image processing of structured light vision technology is to accurately extract the centerline of the laser stripe. [0003] From the analysis of the implementation method, t...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the uneven thickness and curvature changes of the structured light stripes, but the existing methods have low precision, and at the same time, it is difficult to accurately extract the center of the light stripes when the gray scale and width of the light stripes are uneven. In order to solve the position problem, on the basis of considering the normal direction of the light strip, a high-precision sub-pixel extraction method based on adaptive threshold and Fourier fitting is invented.

Images