基于图像处理和轮廓特征点检测的植物叶片几何建模

Geometrical Modeling of leaf Shape Based on Image Processing and Feature Points Detection

摘要: 提出了一种植物叶片轮廓的几何建模方法，对植物叶片的数字图像进行二值化和边缘提取．通过边缘扫描算法，将叶片图像的边缘数据按照像素点原有的拓扑关系生成坐标链表．以坐标链表检测到的轮廓特征点进行三次样条曲线插值，建立植物叶片的几何模型．实验结果表明，该方法较好地保持了植物叶片原有的轮廓特征，并具有数据获取简便、数据量少和几何模型简单的优点．

Abstract: A method of establishing the geometrical model of leaf shape is used. The method starts with the binarization and edge detection of the digital image. According to the topology of the pixels in the edge, a linked list of Coordinates is created by the scanning algorithm. With the feature points detected from the linked list, the geometrical model of leaf shape is upbuilt by the rational cubic spline interpolation. Some experiment results show that this method is more simple and convenient and needs less data.