摘要
针对光源、手背厚度及水纹对手背静脉图像的噪声影响问题,提出一种新的手背静脉特征提取方法,包括最近邻图像插值缩放、Canny边缘检测以及一种改进的基于灰度直方图加权并顾及像素空间信息的FCM算法。实验结果证明,应用改进的FCM算法能减少图像分割迭代次数,较好地滤除噪声干扰。
In order to overcome the effects which light source,the thickness of the hand skin or water refraction take on the hand vein picture,this paper proposes a novel algorithm to extract the hand vein features.It includes the nearest-neighbor interpolation,Canny operator and an improved FCM algorithm considering the information of pixels and based on picture grayscales.Experimental results indicate the new method can reduce the iteration times of image segmentation,and can filter the noise effectively.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第16期154-156,共3页
Computer Engineering
基金
国家自然科学基金资助项目"动态视觉传感器网络若干问题研究"(60973095)
