基于图像处理的叶片主叶脉节点分布信息的提取

Extraction of the nodes' distribution for the main vein based on image processing

植物的叶脉包含着植物重要的生理特征,叶脉及其分布信息的提取是植物分类、识别和建模的基础和关键。利用去叶柄后叶缘的曲率参数函数提取特殊点并利用特殊点确定主叶脉区域,为了避免主叶脉的弯曲和图像局部光照的不均匀对提取效果的影响,对主叶脉区域进行分块,然后利用Hough变换、区域生长和Harris角点检测等算法提取主叶脉区域内主叶脉与次级叶脉之间的节点分布信息,最后汇集节点分布信息得到整片树叶的主叶脉节点分布信息。结果表明,所采用的提取算法可准确地获得节点和节点分布信息。

The vein of plants contains the important physical characteristics of plants. The vine and its distribution are foundation and key to the plants＇ classification,recognition and modeling. After removing the petiole and the special points,which can be used to confirm the region of main vein,are obtained by the Curvature＇s parameter function of the edge. In order to avoid the influence of the bending of the main vein and local illumination uneven on the extraction,the region of the main vein is partitioned. Using the Hough transform,region growing and Harris corner detection,the nodes＇ distribution of the main vein within the partitioned region are obtained,respectively. Then,gathering the nodes＇ distribution of all regions,the nodes＇ distributing information of the main vein is obtained. The results show that the algorithms are capable of accurately extracting the nodes and the nodes＇ distribution.