摘要
提出了一种新的有效的图像阈值分割算法。本算法将小波理论,模糊集理论和信息论三者有机的结合起来。算法采用目标类与背景类最大模糊熵准则,并针对现有的模糊熵分割算法中存在的阈值选取计算效率低的缺陷,重新定义了新的模糊隶属度函数与模糊熵,并且根据小波理论的多尺度分辨的思想,对该算法进行了优化,提高了阈值选取的寻优速度。通过与各种经典算法比较,证实该文算法的可行性,具有很好的鲁棒性。
A new effective image segmentation algorithm Based on Wavelet Operator is presented,which is tied tightly in Wavelet analysis with Fuzzy, information theory.The optimal threshod is searched through a maximal Fuzzy-Entropy criterion,The use of wavelet theory overcomes the problem about the limitation of poorly efficiency, Compared with several traditional algorithms,the merit of the new algorithm is obtained.
出处
《红外技术》
CSCD
北大核心
2004年第3期29-32,共4页
Infrared Technology
