基于区域信息主动轮廓模型的图像分割

Image segmentation based on active contours with local information

图像分割是对图像进行后续处理的关键步骤之一,传统主动轮廓模型在目标图像背景较为复杂的情况下很难精确地进行图像分割。为了精确且快速地进行图像分割,以便更加有利地进行后续相关图像处理操作,在对传统主动轮廓模型进行相关研究的基础之上,提出一种基于区域信息主动轮廓模型的图像分割方法。将图像区域信息融入主动轮廓模型的能量函数中去,减弱了模型对图像区域信息突变所造成的图像误分割;改进该模型能量函数内外曲线的拟合中心,以此减少图像噪声点对拟合中心准确性的影响;利用信息熵改进曲线内外能量函数权重,以此提高曲线的演化速度。实验结果表明,与传统CV(Chan_Vese)模型等四种模型相比,该方法所分割的图像更加精确,且在算法分割效率上具有较明显的优势。

Image segmentation is one of the key steps in the subsequent image processing.When the background of the target is complex,the traditional active contour model is difficult to segment the image accurately.In order to make the image segmentation more accurately,this paper studies the active contour model,afterwards,proposes an image segmentation method based on improved active contour model with local information.Firstly,the local information is integrated into the energy function of the active contour model to enhance the image segmentation caused by the mutation of the local information.Secondly,the fitting centers of the internal and external curves are improved to reduce the effects of image noise on the accuracy of the fitting center.Finally,to speed up the evolution of the curve,the information entropy is used to enhance the weights inside and outside the curve.Compared with the traditional CV model and other three models,the experimental results show that the proposed model can segment image more accurately,and achieve great advantages in the efficiency of segmentation.