Image binarization method based on joint optimization of Otsu method and K-means clustering algorithm

Abstract

The invention discloses an image binarization method based on joint optimization of an Otsu method and a K-means clustering algorithm, belongs to the field of computer vision measurement, and relatesto an image binarization method based on joint optimization of an Otsu method and a K-means clustering algorithm. The method comprises the following steps: firstly, adopting an OTSU algorithm for an image to obtain an OTSU algorithm threshold of the image; obtaining pixel average values of the foreground and the background of the image under the OTSU threshold value respectively; and taking the pixel average value of the foreground and the background as a clustering center of Kmeans, and performing clustering operation on the image, and finally, filtering by adopting a DBSCAN algorithm to obtain a final image binarization result. The method effectively solves the problems that the binarization effect of the image with the local overexposure characteristic in a complex measurement environment is poor, and feature mark points are difficult to extract effectively. Effective and accurate binarization of the image is realized, and the Kmeans clustering process can be completed quickly.

Description

technical field [0001] The invention belongs to the field of computer vision measurement and relates to an image binarization method based on joint optimization of Otsu method and K-means clustering algorithm. Background technique [0002] With the progress of modern industry and the continuous development of computer science, visual measurement technology has been widely used in aerospace, precision manufacturing, product quality inspection and other industrial fields. Among them, the effective binarization of the image of the camera measurement system is the premise and basis for the accurate extraction of feature points, 3D reconstruction, and acquisition of measurement information. However, the industrial measurement site environment is complex, there are various light source interference, and the background of the image is mostly reflective, which makes the image prone to partial overexposure. In the face of the above problems, it is very difficult to accurately realiz...

AI Technical Summary

Problems solved by technology

[0005] An image binarization method based on the combined optimization of the Otsu method and the K-means clustering algorithm of the present invention is to solve the problem that the original image acquired by the camera in the industrial measurement environment is prone to overexposure due to local highlights, and the binarization processing effect is relatively poor. poor this problem

Images