基于形状约束和局部演化的二值水平集运动目标分割.pdfVIP

第35 卷第5 期 电 子 与 信 息 学 报 Vol.35No.5 2013 年5 月 Journal of Electronics Information Technology May 2013 基于形状约束和局部演化的二值水平集运动目标分割 郑 锦 仙 树* 李 波 (北京航空航天大学数字媒体北京市重点实验室 北京 100191) (北京航空航天大学虚拟现实技术与系统国家重点实验室 北京 100191) 摘 要：针对水平集分割模型运算效率较低且易出现过分割的现象，结合视频中运动目标分割的应用背景，该文提 出一种将运动目标检测作为先验形状约束和曲线局部演化方法相结合的二值水平集分割模型。该模型提出将运动目 标检测的区域作为先验形状信息对水平集分割进行约束，并使用二值函数替换传统水平集函数提高运算效率，同时 融入曲线的局部演化方法解决二值水平集模型缺乏曲线演化渐进性的问题。实验结果表明，该文方法在分割准确性、 鲁棒性和运算效率等方面与相关模型相比均有不同程度的提高。 关键词：目标分割；形状约束；局部演化；二值水平集 中图分类号：TN919.8 文献标识码： A 文章编号：1009-5896(2013)05-1037-07 DOI: 10.3724/SP.J.1146.2012.01727 Moving Object Segmentation Using Binary Level Set Based on Shape Constraint and Local Evolution Zheng Jin Xian Shu Li Bo (Beijing Key Laboratory of Digital Media, Beihang University, Beijing 100191, China) (The State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China) Abstract: In order to solve over-segmentation issue and improve computing efficiency, this paper proposes a moving object segmentation model using binary level set based on shape constraint and local curve evolution. Firstly, the model introduces priori shape information in the traditional level set model to constrain segmentation, and the shape is obtained by object detection. Then, to improve efficiency the proposed model uses a binary level set function to replace the traditional level set function. Furthermore, the paper proposes a method of local curve evolution to address the lack of gradual progress in binary level set curve evolution. Finally, the experimental results show that an obvious performance improvement on segmentation