Method and system for level set image segmentation combined with saliency information of brightness correction

Abstract

The disclosure provides a level set image segmentation method and system combined with brightness-corrected saliency information, which performs brightness correction and saliency detection on the target image, and obtains regional saliency information of the corrected image; combined with the energy functional function of the CV model and the energy functional function of the DRLSE model to construct a new energy functional function containing the saliency information of brightness correction; evolve according to the constructed new energy functional function, obtain the curve after the final evolution and perform image segmentation to obtain the segmented result. It guarantees a robust segmentation effect when segmenting images with uneven brightness and complex scene images.

Description

technical field [0001] The present disclosure relates to the technical field of image processing, in particular to a level set image segmentation method and system combined with brightness-corrected saliency information. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] Image segmentation (Image Segmentation) involves various fields of image research, and is one of the very basic and important technologies in image understanding and computer vision. In recent years, the Active Contour Model based on the Level Set Method has become a popular method in image segmentation technology because of its advantages of free topological transformation and strong mathematical foundation. The basic idea is to use the curve evolution theory to embed the curve as a zero-level set on a higher one-dimensional surface, introduce some features of the image to...

AI Technical Summary

Problems solved by technology

[0004] The level set image segmentation method includes two methods based on region and edge. The region-based level set segmentation method mainly uses the regional statistical information of the image. This type of method is less affected by noise and weak edges, but for uneven brightness The image segmentation effect is not ideal; the edge-based level set segmentation method generally uses the gradient information of the image to construct the energy functional, but it is sensitive to the noise in the image, and it is easy to fall into the local minimum and cannot achieve accurate segmentation.

The CV (Chan-Vese) model is a typical region-based level set image segmentation model. Many subsequent region-based level set models are improved on the basis of this model. This model uses the global gray information of the image to construct Energy functional, the biggest shortcoming is that it cannot segment images with uneven brightness well

The DRLSE (DistanceRegularized Level Set Evolution) model is a typical edge-based level set segmentation model. By adding internal energy penalty items, it avoids a series of problems caused by repeated initialization of level set evolution. However, when segmenting images with complex background noise , it is easy for the segmentation curve to deviate from the target area and cause wrong segmentation

Images