An active contour model based on local pre-piecewise fitting image

Abstract

Active contour model (ACM) has been widely used in the field of image segmentation and successfully solved numerous problems. Segmentation of intensity inhomogeneity images is a major focus of current research. To seek a better solution, this paper proposes an active contour model based on local pre-piecewise fitting image. First of all, the piecewise function is pre-fitted by the dilation and erosion operator in gray morphology. Second, a novel energy function is drawn out with a rewritten image fitting function. Third, an optimized gradient descent flow formula is proposed to further improve the robustness of this model. Fourth, an improved regularization term is proposed to optimize the level set function. Experiments have proved that the model can effectively segment images with uneven grayscale. Compared with other models, the proposed model has obvious advantages in robustness to initialization and segmentation efficiency. In addition, the model can also choose various targets by selecting different initial contours.