Research Of Image Segmentation Based On The Minimum Spanning Tree

Image segmentation is the process of segmenting an image into regions with various characteristics and proposing the target region.It is a key step in subsequent image processing,and it is also the basis for image understanding and advanced applications.Image segmentation technology has been applied to various fields such as: biomedicine,military public security,industrial engineering and other fields,and has achieved good results.In the field of image segmentation,because there is a good one-to-one correspondence between image and graph,the problem of image segmentation is transformed into the problem of graph segmentation,and the theoretical knowledge of graph theory is used for image segmentation,which can also avoid the error caused by image discretization.Image segmentation methods based on graph theory are popular because of their simplicity and efficiency.In this paper,the image segmentation method based on graph theory is studied.The main work includes the following aspects:1.This article reviewed the basic knowledge of graph theory,and expounds the relation between the graph and image,on the basis of analyzing the model of image segmentation based on graph theory algorithm theory,including the model of image segmentation based on minimum spanning tree,the model of image segmentation based on the shortest path,the model of image segmentation based on cut set standards,model of image segmentation based on graph cut theory,and discuss the principle and performance of the typical algorithm.2.In the image segmentation model based on graph theory,we focused on the analysis and study of the image segmentation model method based on the minimum spanning tree,and discussed the segmentation algorithms that currently exist in this field.And point out the shortcomings of the weight judgment standard of the segmentation model in image segmentation.It caused the segmentation results obtained by the segmentation model to often appear under-segmentation and over-segmentation,resulting in low segmentation accuracy,and the processing of image details cannot achieve the desired result.To tackle this challenge,this paper selects a new weight judgment standard,which not only considers the sensitivity of human eyes to the change of RGB value,but also takes into account the spatial distance and vector relationship between two pixels.Therefore,combined with the image segmentation model of minimum spanning tree,this paper proposes a new image segmentation algorithm based on minimum spanning tree.3.Through experimental analysis,the image segmentation model based on the minimum spanning tree proposed in this paper weakens the under-segmentation and oversegmentation of previous image segmentation,improves the accuracy of image segmentation,and the method has good robustness against noise.It can better retain the detailed features of the image and ensure the consistency of the region.At the same time,the segmentation method is applied to the field of medical imaging,and good results have also been achieved.The experiments have proved the effectiveness of the method,and has certain practical application value.However,when the method is used to process complex images,the segmentation time will be too long and the wrong boundary will be produced,which needs further improvement.