A method and system for image segmentation

Abstract

The invention provides an image segmentation method and a system. According to the image segmentation method, an original image is segmented into multiple pixel blocks, each pixel block is taken as a pixel point and is mapped into a communication graph, the quantity of vertexes and edges of the communication graph is greatly reduced, the image segmentation consumption time is shortened, an image segmentation speed is improved, a smallest generation tree of the communication graph is acquired, each node of the smallest generation tree corresponds to a pixel block of the original image, a sampling image segmentation process utilizing the smallest generation tree is actually a process for detecting how the original image is segmented into each small zone, compared with a segmentation method in the prior art simply utilizing color difference of each pixel of an original image to realize segmentation of each pixel point, an original image segmentation speed is improved, image segmentation precision is improved, and a segmentation result of the original image is more consistent with understanding of a mankind visual sense system about perception objects.

Description

technical field [0001] This application mainly relates to the field of image processing, and more specifically relates to an image segmentation method and system. Background technique [0002] At present, image processing is widely used in medical images, remote sensing cloud images, fingerprint recognition, face detection, geological survey and other fields. As a key step in image processing, image segmentation is used to provide effective information for image retrieval and image analysis. This makes higher-level image processing possible. [0003] Among them, image segmentation is to use certain characteristics of the image, such as grayscale, color, texture, etc., to divide the image into several independent and meaningfully connected regions or objects, each region has the same characteristics, through these regions To express the designed scene or object, conform to the visual characteristics of the human eye in reality, and meet the user's image processing needs. ...

AI Technical Summary

Problems solved by technology

[0004] The commonly used image segmentation methods in the prior art mainly include edge detection-based or region-based image segmentation methods. Although the purpose of image segmentation can be achieved, the original image segmentation effect of these existing image segmentation methods is not ideal, often Does not match the user's visual system's understanding of perceived objects

Images