Phase extraction-based no-color difference three-dimensional character image processing method

Abstract

The invention discloses a phase extraction-based no-color difference three-dimensional character image processing method. The method comprises steps: image acquisition is carried out, a laser scanning galvanometer is used for reflecting in a time sharing mode laser strips transmitted by a single-line laser on the surface of the three-dimensional character, a camera is then used for acquiring distorted light strip images, the distorted light strip images are stacked, and a raster image with a three-dimensional character topological structure is obtained; phase extraction is carried out on the obtained raster image on the basis of fourier transform, phase unwrapping is carried out on the extracted phase, an absolute phase of the three-dimensional character is obtained, absolute phase calculation is carried out on the background, the relative phase of the three-dimensional character is obtained according to the absolute phase of the three-dimensional character and the absolute phase of the background, the relative phase replaces the gray value of the character, and the gray image of the three-dimensional character is obtained through reconstruction. Compared with the prior method, stability is higher, the extraction process is simpler, the efficiency is higher, and engineering application is more easily realized.

Description

technical field [0001] The invention relates to a method for processing a three-dimensional character image without color difference based on phase extraction. Background technique [0002] Three-dimensional characters without color difference are widely used in the fields of industrial production and logistics. The outstanding features of this type of characters are "no color difference" and "three-dimensional". Characters are the same color as the background, but the heights of the characters and the background are different, which are three-dimensional characters. figure 1 It is an application example of three-dimensional characters without color difference, and the three-dimensional characters without color difference are inside the frame line in the figure. [0003] Because the characters and the background have the same color, the effect of directly collecting character images with a camera is not ideal, and the character recognition effect is poor. For this reason, p...

AI Technical Summary

Problems solved by technology

[0004] 1. The uneven reflection of the laser beam on the line, or the measured surface with stains, pits, etc., the gray scale of the collected distorted light strips is uneven, which makes it difficult to extract the center line, or the center line is broken, and there are multiple center lines locally, etc. If the situation occurs, the subsequent distortion-grayscale transformation cannot be performed, which makes the character image acquisition effect unsatisfactory;

[0005] 2. When the environment changes, such as changes in reflectivity due to different surface materials to be tested, or changes in background light intensity and camera parameters, the quality of the obtained distorted light strip image is unstable, and the light strip centerline extraction algorithm has no effect on the distortion. The light bar image quality is very sensitive, so the existing image processing methods are not stable in practical engineering applications;

[0006] 3. Acquiring a line of character images requires processing multiple light strip images, the algorithm is complex and the amount of calculation is large

Images