燕山大学学报Issue(4)：353-357,5.DOI:10.3969/j.issn.1007-791X.2013.04.010
基于流场纹理表达与表面粗糙度测量的运动目标分割
Moving target segmentation based on flow field texture representation and surface roughness measurement
摘要
Abstract
Aiming at the drawback of region impairment and inside holes detected by background subtraction and optical flow based method, a moving target segmentation method based on flow field representation and surface roughness measurement is proposed. The motions of targets are described by the flow of fluid by computing the optical flow between two consecutive frames. A Line integral convolution method is used to represent the textures to reveal more details of flow field. Therefore, the moving targets and background regions are expressed by different texture images. Finally, the textures of moving target and background regions are mapped as objects with different surface roughness by the strategy of surface roughness measurement with stylus probe method. Furthermore, the threshold for segmentation the surface roughness image is gained by analyzing the histogram. Experiment results show that the proposed method can select the threshold adaptively, and can overcome the problem of inside holes.关键词
运动目标分割/光流场/线积分卷积/表面粗糙度测量/针描法Key words
moving target segmentation/optical flow field/line integral convolution/surface roughness measurement/stylus probe method分类
信息技术与安全科学引用本文复制引用
张旭光,贺海明,苏连成,李小俚..基于流场纹理表达与表面粗糙度测量的运动目标分割[J].燕山大学学报,2013,(4):353-357,5.基金项目
国家自然科学基金资助项目(61271409，60905046)；河北省自然科学基金资助项目(F2013203364)；中国博士后科学基金资助课题(2012M510768，2013T60264)；中国留学基金资助课题(2011813018) （61271409，60905046）
