一种快速提取植物叶片最小外接矩形的算法

Fast Algorithm for Extracting Minimum Enclosing Rectangle of Plant Leaves

为了提高提取植物叶片最小外接矩形的计算效率与精确度,提出一种快速提取植物叶片最小外接矩形的算法。该算法首先使用Canny算子提取叶片轮廓,然后使用基于平面扫描法的Graham算法构造叶片轮廓凸包,最后提取叶片最小外接矩形。仿真实验结果表明:在Flavia植物叶片数据库中进行测试,该算法优于旋转法、顶点链码法。

In order to improve the extraction efficiency and accuracy,a fast algorithm is proposed to extract the minimum enclosing rectangle （MER） of plant leaves. Using the canny edge detection operator to extract the contour of leaves,the Graham algorithm based on the fiat plane scanning is applied to construct the convex hull of the leaf contour,the MER is extracted. The effectiveness of the proposed algorithm is verified by using the Flavia plant leaf database and the experimental results show that it is better than the rotation algorithm and vertex chain code algorithm.