A threshold segmentation algorithm based on neighbourhood characteristics

摘要

Image threshold segmentation, which means to divide the image data into different regions by giving the appropriate segmentation threshold and the corresponding based on the characteristics of the image and the corresponding rules, is a widely used and effective method in the image processing. This article, which introduces the classic method of image segmentation principle and their advantages and disadvantages firstly, investigates image neighbourhood gray features basic theories, and analyses the visual characteristics segmentation images of practical engineering application. It mainly presents a new image visual characteristic test operator used for the project image object detection and segmentation system. This algorithm can obtain good segmentation effect through the contrast.