摘要
针对手背静脉图像的结构和特点,提出基于曲波变换的分水岭算法,对手背静脉图像进行骨架特征提取。文中对远红外手背静脉图像进行分析。实验结果表明,该算法能够有效地提取手背静脉骨架特征,在保持边缘信息的同时,大大减少局部最小值区域的数量,解决分水岭的过分割现象。同时该算法也为静脉骨架特征提取提供了新的途径。
Aimed at the structure and features of the hand vein image, a watershed algorithm based on curvelet transform is proposed to solve the extraction problem. Far infrared images of dorsal hand vein are analyzed. Experimental results show that the algorithm can effectively ex- tract the skeleton and preserve the edge of vein images and restrain the over-segmentation phe- nomenon in the watershed at the same time. Finally, the algorithm provides a new approach for the extraction of the vein image skeleton pattern.
出处
《数据采集与处理》
CSCD
北大核心
2009年第5期655-659,共5页
Journal of Data Acquisition and Processing
关键词
静脉图像
骨架
分水岭算法
曲波变换
vein images
skeleton
watershed algorithm
eurvelet transform
