基于DSP系统的静脉识别算法研究  

Research on Vein Recognition Algorithm Based on DSP System

作　　者：宦键[1,2] 刘国海[1] 梅从立[1] 张东林[2] 陈玉[1] 

机构地区：[1]江苏大学电气信息工程学院,江苏镇江212013 [2]江苏科技大学电子信息工程学院,江苏镇江212003

出　　处：《现代科学仪器》2014年第4期42-45,共4页Modern Scientific Instruments

摘　　要：随着静脉识别技术研究深入,手指静脉识别技术已逐步走向产品化的道路。考虑其应用场合,我们要求生产出来的产品有较高的识别精度,较快的识别速度,还要便携。为此,我们采用ADSP-BF561为核心处理器设计识别系统硬件,提出了一种基于高斯高通滤波器的双阈值静脉分割算法。该方法利用改进的高斯高通滤波器对处理后的图像进行滤波,并对滤波后的图像采用双阈值分割算法确定手指静脉区域。通过实验表明,该方法较单一阈值方法能更精准的实现静脉图像分割。With the vein recognition technology studies in depth,fi nger vein recognition technology has been gradually moving towards the road of product.Taking its applications into consideration,we require its products owning higher recognition accuracy,faster recognition speed and portable.Therefore,we used ADSP-561 as its core processer to design recognition system.We proposed a double image segmentation threshold algorithm based on high-throughput Gaussian filter.The method used an improved high-throughput Gaussian filter to filter the processed image.Then,we segmented the image by using the double image segmentation threshold algorithm to determine the fi nger vein region.Experiments showed that the method is more accurate than a single threshold to achieve image segmentation.

关 键 词：静脉识别 DSP 高斯高通滤波器 双阈值分割算法 

分 类 号：TN391[电子电信—物理电子学]