农机化研究Issue(6)：41-44,4.
基于分块阈值和边缘检测的叶片分割算法
Leaves Segmentation Algorithm Based on Block Threshold and Edge Detection
摘要
Abstract
For complex plant leaf images , a segmentation algorithm was presented by combination of sub-block threshold and edge detection.First, according to a pre-determined sub-block size, the entire image was divided into a certain number of sub-blocks and each sub-block was segmented using the Otsu method , the sub-image after segmentation spliced together to form a target image .Then , the original image edges were extracted by improved sobel operator .The two above segmentation results of original image were combined together for better results , finally morphological operations of corrosion and fill holes were applied for the final segmentation result .Experimental results show that comparing with tra-ditional block threshold , edge detection our algorithm not only has better characteristic of noise resistance and show more clearly details of segmentation results , but also have a better segmentation results .关键词
图像分割/叶片/分块阈值/边缘检测/形态学Key words
image segmentation/leaves/block threshold/edge detection/morphologic分类
信息技术与安全科学引用本文复制引用
赵斌,周厚奎,冯海林..基于分块阈值和边缘检测的叶片分割算法[J].农机化研究,2014,(6):41-44,4.基金项目
国家自然科学基金项目（61272313）；浙江省科技厅公益性项目（2012C21015）；浙江省教育厅项目（Y201224063）；浙江农林大学校发基金项目（2006FR029）；浙江农林大学信息工程学院创新训练计划项目 （）
