A Research on Eztracting Low Quality Human Finger Vein Pattern Characteristics
Based on fuzzy theory and combined the image segmentation algorithm and concave detection algorithm, this paper proposed an enhanced method for extracting finger vein features. Firstly, based on the concave configuration characteristic of the finger vein image, the detection operator is defined to detect the concave area. Secondly, the finger vein image is segmented three regions through by twice coarse segmentation. The three regions are objects region, background region and fuzzy region. The fuzzy region is improved by enhancement algorithm then. Finally, the enhancement region is subdivided. The experiment results show that it is not only effective in removing noise and false vein feature, but also extracted vein features is smooth and continuous. So it is very perfect to extract features of low quality human finger vein image.
fuzzy theory image segmentation concave detection finger vein image feature eztraction
Yu Chengbo Qing Huafeng Zhang Lian
Research Institute of Remote Test & Control, Chongqing Institute of Technology, Chongqing, China
国际会议
上海
英文
1876-1879
2008-05-16（万方平台首次上网日期，不代表论文的发表时间）