摘要
指纹图像的采集处于指纹识别系统的最前端,图像质量的好坏直接关系到系统的处理速度和准确性。针对目前广泛使用的半导体指纹传感器,根据指纹本身的固有特性,设计出一组多通道Gabor滤波器对指纹图像进行分解和特征提取,并结合基本的图像质量评估指标进行质量分析和判断。实验结果表明:该算法能有效地判断半导体指纹传感器的图像质量,有助于提高自动指纹识别系统的整体性能。
As the forefront of the fingerprint identification system, the quality of fingerprint image determines its speed and precision. Based on the inherent feature of fingerprint and the basic image quality estimating methods, a new estimating approach for fingerprint image quality captured by semiconductor sensors is proposed in this paper. Multi-channel Gabor filters are designed for image analysis and feature extraction. Experimental results show that this algorithm is quite effective in fingerprint image quality estimation, and help to improve the performance of the fingerprint identification system.
出处
《光电工程》
EI
CAS
CSCD
北大核心
2006年第10期104-107,144,共5页
Opto-Electronic Engineering
基金
重庆市科技攻关项目资助项目(CSTC.2004AA-2001-8277-02)
