基于霍夫变换与角点检测的叶脉特征提取算法

A Feature Extraction Algorithm for Leaf Vein Based on Hough Transform and Corner Detection

叶脉特征提取对植物的基因研究具有重大价值,叶脉的特征体现了叶片的各种信息。而叶脉的角度信息直接反映了叶脉生长的方向和叶片顶端、底端的生长形态。因此,为了克服现有算法的局限性,提出一种基于霍夫变换和角点检测的叶脉特征提取算法。该算法首先通过霍夫变换搜索主叶脉,通过筛选得到主叶脉大致所在直线,继而得到叶片顶端、底端坐标;然后通过角点检测的方法搜索主叶脉与次叶脉的交点,再以获得的交点为起点,以区域增长法搜索叶片,以进一步提取主叶脉和次叶脉的坐标等参量,同时测得角度特征,包括主叶脉与边界夹角和主次叶脉夹角。利用瑞典白杨树叶(12000片)对该算法进行验证,结果表明该算法在主次叶脉提取及其角度检测方面效果良好。

The extraction of leaf vein features is of great value to the genetic research of plants.The characteristics of leaf veins reflect various information of leaf.The angle of leaf vein directly reflects the direction of vein growth and the shape of leaf top and bottom.Therefore,to overcome the drawbacks of current veins extraction algorithms,a novel feature extraction algorithm for leaf vein based on Hough transform is presented.Firstly,the main vein is searched by Hough transform,and the approximate line of the main vein is obtained by filtering,and then the coordinates of the top and bottom of the leaf are obtained.Secondly,the intersection point between the main vein and the secondary vein is searched by corner detection method,and the area growth method is used to search the leaves from the intersection point obtained as the starting point,so as to further extract the coordinates of main and secondary veins and other parameters.Thereby,the angle parameters are calculated,including the angle between main and secondary vein,and that between the primary vein and leaf boundary.The algorithm is verified with leaves of Swedish aspen(12000 leaves),which shows that it is effective in the main and secondary vein extraction and angle detection.