基于高光谱图像技术的指纹识别研究  

Study on Detection of fingerprint recognition Based on Hyperspectral imaging technology

作　　者：赵明富[1] 夏曦[1] 张政委[1] 冯小平[1] 

机构地区：[1]重庆理工大学电子信息与自动化学院,重庆400054

出　　处：《激光杂志》2013年第1期45-47,共3页Laser Journal

摘　　要：目前,国内外通过高光谱图像技术来进行指纹识别的研究非常少。高光谱图像技术作为集图像信息与光谱信息于一身的新技术,对指纹识别的可行性做了探讨和研究。以血指纹为目标,首先对高光谱指纹图像进行了主成分分析,优选出415nm、550nm和630nm三个特征波长进行图像融合;然后对该图像进行中值滤波得到更为清晰指纹图像;最后通过MATLAB软件对指纹某个像素点的光吸收强度值进行提取,得到该像素点的光谱曲线。实验研究表明,用高光谱图像技术来识别指纹是可行的。At present, it is very little by hyperspectral imaging technique to fingerprint identification at home and abroad. The new technology of hyperspectral imaging technique that it is include the image information and spectral information , Explore and study the feasibility of fingerprint identification. Target blood fingerprint, First research component analysis of high spectral fingerprint image principal, Preferred Three characteristic wavelength for 415.550and630 Nano to Image Fusion. Median filter and then get the clearer fingerprint image. Finally, through MATLAB software, the fingerprint of a pixel point light absorption intensity values were extracted, get the spectral curve of the pixel point. Experimental studies have shown, Hyperspectral imaging technique to identify the fingerprint is feasible.

关 键 词：高光谱图像技术 主成分分析 指纹 谱曲线 

分 类 号：TN248.1[电子电信—物理电子学]