基于Canny算子的指静脉图像特征提取算法研究

Research on Image Feature Extraction Algorithm of Finger-vein based on Canny Operator

主要对指静脉图像的特征提取方法进行了研究。首先用双阈值Canny算子提取指静脉边缘轮廓,并结合医学测量数据确定了指静脉图像的ROI区域;然后利用灰度和距离的相关性解决了伪边缘信息的剔除和不连续边缘的修复;最后对修复后的边缘信息进行了形态学操作处理。模拟实验结果表明,本文算法简单有效,提取的指静脉边缘特征信息清晰完整,为识别奠定了较好的基础。

The feature extraction method of finger-vein image is studied in this paper.Firstly,the double threshold Canny operator is used to extract the edge contour of digital vein,and the ROI region of finger-vein image is deter⁃mined with medical measurement data.Then,the correlation between gray level and distance are used to eliminate the false edge information and repair the discontinuous edge.Finally,the repaired edge information is processed by morphological operation.The simulation results show that the algorithm is simple and effective,and the extracted fin⁃ger-vein edge feature information is clear and complete,which lays a good foundation for recognition.