摘要
通过分析播种机性能检测的方法和种子目标的图像特点,提出一种基于数学形态学的种子图像边缘检测方法。利用开闭运算的级联对种子图像在生成过程中因传送带上的杂质等的干扰进行滤波后,采用3×3八方位方形结构元边缘检测算法获取种子的边缘。试验表明,与经典的边缘检测算子相比,该算法不仅具有很好的边缘提取能力,而且具有很强的抗噪能力。
In this paper, by analyzing the methods of seedmeter quality detection and the features of seed image , a new seed edge measuring method based on mathematical morphology is introduced. First , the noise of seed edge image was eliminated by use of morphological opening and closing algorithm ,which is caused by the impurity on the belt, vibration and so on.after that, seed edge is detected by 3×3 multi direction morphological structuring elements edge detection algorithm. The examples show that the method can obtain better quality of edge detection than the classic operators of edge detection.
出处
《湖北农业科学》
北大核心
2008年第5期587-589,共3页
Hubei Agricultural Sciences
基金
河南省教育厅自然科学研究计划项目(2007210005)
关键词
排种器
种子边缘
数学形态学
多方位
seedmeter
seed edge
mathematical morphology
multi direction
