关键词：
枣树叶片;田间图像;阈值分割;Canny算子;形态学
摘 要：
设计了一种复杂图像分割的综合算法。首先对图像进行预处理,锁定大致目标范围,对图像进行增强,再基于阈值分割和Canny算子对图像进行初步分割。然后结合形态学处理方法及各种逻辑运算对分割结果进行优化处理,得到精确完整的目标图像。以田间枣树叶片图像为例进行实验,证明了该算法的可行性和有效性。该算法对叶片重叠、叶片灰度不均匀等复杂图像都有很好的分割效果,获得了边缘清晰、平滑、定位精确的边缘图像。
译 名：
Complex Target Image of Field Jujube Leaf Segmentation Based on Integrated Technology
作 者：
Dong Jinyong Wang Jianlun Li Daoliang He Jianlei Wang Yongbin (College of Information and Electrical Engineering,China Agricultural University,Beijing 100083,China)
关键词：
Jujube leaf,Field image,Threshold segmentation,Canny operator,Morphological methods
摘 要：
To segment the complex target image,an algorithm based on integrated technology was proposed.First,the image was preprocessed to find the target area and have the image enhanced.Second,the target area was segmented by threshold and Canny operator,some raw results were acquired.Finally,in order to get the precise and complete target image,the optimizing process of the algorithm was made with morphological methods and logical calculation.Experimental results of field jujube leaf image showed that this algorithm was feasible and effective.This algorithm had a great advantage in complex leaf image which contains overlap leaves and uneven gray scale.The clear,smooth,precise edge image could be obtained.
