Relative-Target-Complexity-Based Clutter Metric for Human Visual Systems

摘要：

Under the assumption that the image information is determined mostly by the scene illumination and the objects structure, a new clutter metric, named the relative target complexity based clutter metric, for human visual systems is presented in this paper. The average intensity and the target complexity, independent of each other, are used to represent the image signal in the new metric. And its promise is demonstrated by the experiment results. Furthermore, the comparison, based on the search_2 database, of the new metric to the classical clutter metrics such as POE and SV shows its superiority.

作者: Qian Li Cui Yang Jian-Qi Zhang

作者单位: School of Technical Physics, Xidian University, Xian Shaanxi 710071, China

会议类型: 国际会议

会议名称: 2010 International Conference on Audio,Language and Image Processing(2010年音频、语言与图像处理国际会议 ICALIP 2010)

会议地点: 上海

会议语种:英文

页码: 1195-1199

在线出版日期: 2010-11-23（万方平台首次上网日期，不代表论文的发表时间）