东南大学学报（自然科学版）Issue(1)：35-41,7.DOI:10.3969/j.issn.1001-0505.2016.01.007
一种基于近红外可见光映射的快捷静脉提取算法
A fast vein detection algorithm based on near infrared-visible light mapping
摘要
Abstract
A fast vein detection algorithm is proposed for visible images.Pairs of skin images are captured by using a visible-near infrared ( NIR) synchronized camera.The contrast of the NIR ima-ges is adjusted by the contrast limited adaptive histogram equalization method.Gabor filters are adopted to automatically detect veins information.Based on it, the corresponding pixel values of veins and general skin are exacted from the visible-NIR images and used as samples to train a three-layered feed-forward neural network, which approximates the mapping from RGB ( red, green and blue) values to NIR intensities.Veins can be observed from the mapped images.To eliminate the influence of illumination variations on pixel values in visible images, an image intensity adjustment algorithm is proposed.The experimental results show that the proposed algorithm is robust to images taken from different people under different illumination and camera conditions.It also has a stable performance on images taken at different times.The computation is simple and fast, and it is helpful to develop low-cost portable vein imaging device for medical treatment.关键词
静脉提取/可见光图像/近红外图像/神经网络/Gabor滤波器Key words
vein detection/visible image/near infrared image/neural network/Gabor filter分类
信息技术与安全科学引用本文复制引用
唐超颖..一种基于近红外可见光映射的快捷静脉提取算法[J].东南大学学报（自然科学版）,2016,(1):35-41,7.基金项目
国家自然科学基金资助项目（61403196）、教育部博士点基金资助项目（20133218120018）、江苏省自然科学基金资助项目（BK20140837）． （）
