一种新的基于局部区域的活动轮廓模型

摘要

提出了一种新的基于曲线演化的活动轮廓图像分割模型.该模型利用局部图像统计信息来代替C-V模型中的全局灰度均值,以此可以分割灰度不均匀的图像.此外,在模型定义的能量泛函中增加了水平集正则项,以此来保证数值计算的准确性和避免对水平集函数的重新初始化.将本文提出的活动轮廓模型用于分割人工和自然图像,比较结果显示：C-V模型不能很好处理灰度不均匀图像,而本文提出的模型对灰度不均匀图像能得到满意的分割效果.%A novel active contours model for image segmentation based on techniques of curve evolution is proposed.The proposed model utilizes the local image information to replace the global average intensity of C-V model,it is able to segment images with intensity inhomogeneities.Moreover,it adds to the level set regularization term in the energy functional to ensure accurate computation and avoids expensive re-initialization of the level set function.The proposed model is applied to synthetic and real images,and the results show that the C-V model can not deal with heterogeneous images,but,the proposed model can do well.