基于Gabor滤波和形态学变换的手臂静脉线提取

摘要

An algorithm of arm vein extraction for near-infrared images is presented.Firstly,the contrast limited adaptive histogram equalization(CLAHE)is used to enhance the contrast of near-infrared image.The Gabor filters are uti-lized to the processed image to obtain the orientation and energy images,which are used to acquire enhanced arm vein image. The top and bottom hat transformation are used on enhanced image.Then the whole image is transformed to binary image. The morphological open and close operation are applied on the binary image.The arm vein lines are extracted by using the morphological skeleton method.At last the more complete and effective vein line are obtained after burr trimming of the skeleton image.The experimental results show that the proposed algorithm can effectively guarantee the continuity and inte-grality of vein lines,and improve the matching accuracy of the arm vein efficiently.%提出了一种针对近红外图像的手臂静脉提取算法.首先利用限制对比度自适应直方图均衡化(CLAHE)对近红外图像进行预处理,增强对比度.然后利用Gabor滤波器对增强后的图像进行滤波,由Gabor滤波后的方向图和能量图获得手臂静脉位置.接着,对能量图进行高低帽变换,然后将整幅图像进行二值化,采用开闭运算处理二值化图像,利用形态学骨架提取的方法提取手臂静脉线,最后对骨架化后的图像进行毛刺修剪,得到比较完整有效的静脉线.实验结果表明,本文提出的算法可以有效地保证静脉线的连续性和完整性,有效地提高手臂静脉的匹配率.