Improvement Of CV Model And Its Research In Image Segmentation Application

Image segmentation is the basis of image processing technology,and it has a wide range of application in real life.Among the many methods of image segmentation,CV models based on the level set method have attracted much attention from scholars.Although the method has achieved certain results,there are still some problems.For example,the segmentation result depends on the selection of the initial contour,and the result of image segmentation with uneven gray scale distribution is not ideal.Based on this defect,this paper improves the traditional CV model and applies it to the study of wood image defects.The specific research contents are as follows:Aiming at the shortcomings of the CV model for the image segmentation with blurred boundary features and background gray unevenness,The boundary indication function is modified,and the improved boundary indication function is integrated into the length term of the CV model.Refer to the distance regular term of the double well potential to avoid re-initialization of the level set,and a level set evolution equation combining gradient and region information is obtained.Functional theory is applied to prove that there is a unique solution to the level set equation,and the equation is solved by the finite difference method in the variational method.To overcome the shortcoming of CV model in image segmentation with high noise and uneven background gray level,The global information of CV model is combined with the local information of LBF model.In order to reduce the influence of the noise point on the precision of the fitting center of the traditional CV model and the accuracy of the fitting center of the LBF model,the median is used instead of the mean to improve the fitting center.In order to effectively adjust the weights between global and local terms,this paper establishes a weighting function based on the gray mean difference of small areas inside and outside the curve contour,which improves the efficiency and accuracy of the model to segmentation of wood defects to verify its effectiveness.