摘要
玉米籽粒形状特征是玉米品质和品种判定的重要依据,利用机器视觉代替人的视觉对玉米籽粒品种或品质的自动检测是发展的必然趋势。为此,通过图像获取、颜色空间转换、图像分割、形态学处理、单籽粒图像的定位和拾取、特征参数提取等步骤实现了对玉米籽粒图像的特征提取。所采用的算法简洁、有效,对下一步玉米籽粒的自动检测工作具有重要的实际意义。
The shape feature of maize is one of the important aspects of the quality of corn. Automatic detection using machine vision instead of human vision on maize seed quality is the inevitable trend of development. In this paper, we did image feature extraction on maize seeds by the processes of image acquisition, image color space conversion, image segmentation, morphological processing and characteristic parameters extraction. The algorithm we used is concise and effective. It is significance to the next automatic detection work of maize seeds.
出处
《农机化研究》
北大核心
2015年第6期46-48,52,共4页
Journal of Agricultural Mechanization Research
基金
吉林省教育厅"十二五"规划项目(201356)
关键词
玉米籽粒
形状特征
图像处理
品种检测
maize seeds
physical properties
image processing
quality inspection
