摘要
为了更有效将Tchebichef矩应用于模式识别和图像分析等领域,提出一种关于Tchebichef矩的快速有效算法.对二值图像采用图像块描述方法,并在推导了Tchebichef多项式一些性质的基础上,实现了一种快速计算Tchebichef矩的新方法.实验结果表明,该方法能够有效地节省计算时间,尤其在图像尺寸较大,以及所使用矩的阶数较高的情况下,该算法的效果更为明显.
To expand the application of the Tchebichef moments in the fields of pattern recognition and image analysis, a novel algorithm for fast computation of Tchebichef moments is presented. By using the image block representation to describe a binary image, and by establishing some useful properties of Tchebichef polynomials, an efficient method is proposed for computation of Tchebichef moments. Experimental results show that this method can save much CPU (central processing unit) time in the calculation of moments, especially in the case where the image size is large and the high order of moments is used.
出处
《东南大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2006年第5期857-862,共6页
Journal of Southeast University:Natural Science Edition
基金
科学技术部基础研究重大资助项目(2003CB716102)
国家自然科学基金资助项目(60272045)
教育部新世纪优秀人才支持计划资助项目(NCET-04-0477)