摘要
基于对传统二维最大熵算法各种改进方法分析,提出了一种基于判决域自动约束的二维最大熵改进算法。该算法提出了可以自动确定判决域大小的经验公式,与以往的改进算法相比,不仅减少了算法的运算量,同时具有自适应性,因而在某些应用场合具有较强的实用性。
A modified algorithm of two-dimensional maximum entropy thresholding segmentation, which is based on Verdict Scope Adaptive Restriction(VSAR), is proposed in this paper by analyzing modified methods of two-dimensional maximum entropy algorithm. The new modified algorithm puts forward a experienced formula which adaptively determines the verdict scope. Compared with previous modified algorithms, the proposed modified algorithm not only reduces the running time, but also bears the adaptivity in some way, thus it possesses practicability in some applications.
出处
《计算机应用研究》
CSCD
北大核心
2004年第12期29-30,42,共3页
Application Research of Computers
基金
国家"973"计划资助项目(2002CB312200)
关键词
阈值分割
最大熵
二维直方图
Thresholding Segmentation
Maximum Entropy
Two-Dimensional Histogram
