Research Of The Geometric Active Contour Model In Image Local Segmentation With Homogeneous Intensity

With the rapid development of electronic,computer technology and image acquisition technology,the application of image processing technology has received great attention.Image segmentation is a key technique in image processing.The aim is to divide the image into regions with different characteristics and extract the target of interest.The energy-based segmentation method is an algorithm developed on the basis of the active contour model.By defining an energy function which variables include edge curves and regional data items,and when the functional energy reaches the maximum,the curve position is the contour of the target.The goal of this paper is to study geometric active contour model in image local segmentation with homogeneous intensity by using level set method.To the problem that the energy function models of global energy and local energy both are unable to quickly and accurately segment the image with homogeneous intensity,a local segmentation algorithm is proposed,which target energy function is constructed by the adaptive coefficients with the global item(CV model)and the local item(LGDF model).In the early part of the algorithm,The global item plays a leading role,leading to the target contour curve near.When the evolution curve is close to target outline,local item plays a leading role,leading the evolution curve further close to the target contour.The mean value of the absolute value of the difference between the gray mean of the local neighborhood of the points on the evolution curve and the mean value of the whole inner gray scale of the evolution curve is taken as the weight coefficient,and the adaptive adjustment between the global term and the local term can be realized to image segmentation.Aiming at the problem of narrowband control instability and accuracy lack of curve evolution in classical gradient-based and curvature-based narrow band construction methods,a method of constructing narrow band based on particle swarm threshold is proposed.The method can control the width of the narrow band by changing the radius of the morphological operator,not only improve the stability of the narrow band,but also optimize the energy calculation area of the curve evolution,which greatly improves the evolution precision and efficiency.