Tsallis熵和改进CV模型的海面溢油SAR图像分割

Marine spill oil SAR image segmentation based on Tsallis entropyand improved Chan Vese model

为了解决海洋表面溢油监测中合成孔径雷达(SAR)图像分割精度不高的难题,提出一种基于Tsallis熵多阈值分割与改进CV(ChanVese)模型相结合的海面溢油图像分割方法。首先采用基于Tsallis熵的多阈值选取算法对海面溢油图像进行粗分割;然后分别将得到的溢油区域和溢油粗略轮廓作为CV模型的局部区域和初始轮廓,以降低CV模型的场景复杂度及其对初始条件的敏感性。CV模型仅考虑了图像各区域的均值信息而没有考虑图像的局部信息,尽管能够得到渐进型边界图像,但其分割结果存在误差。本文采用了加入移动因子的改进CV模型降低分割误差,提高收敛速度。实验结果表明,提出的海面溢油SAR图像分割方法具有分割边界定位准确、运行高效和无需设置初始条件等优点。

Considering the low accuracy of Synthetic Aperture Radar （SAR） image segmentation in the marine spill oil detection, a segmentation method of marine spill oil images based on Tsallis entropy multilevel thresholding and improved Chan Vese （CV） model is proposed in this paper. First, the multi-threshold selection algorithm based on Tsallis entropy is used to make a coarse segmentation for marine spill oil images. The obtained spill oil region and its coarse contour provide local region and initial con- tour for CV model, respectively, which are used to reduce the scene complexity of CV model and its sensitivity to initial situation. The traditional CV model only considers the mean value of each region of image instead of the local information of image. Though it can get non-gradient defined image boundary, there are errors in the segmented results. We use an improved CV model with the motion factor, thus the segmentation errors are reduced and the convergence speed is increased. Experimental results show that the our method not only dispenses with initial condition, but also ensures accurate segmentation boundary and efficient operation.