摘要
在对955个玉米颗粒图像研究的基础上,探讨了用自然光和白炽灯照射下获取的RGB彩色模型来识别、分割玉米图像的可行性。利用图像识别软件,提取了玉米颗粒的面积、周长、直径、圆形度和延伸度等几何特征;通过对图像特征参数的分析,提出用直径和圆形度参数来表征玉米的几何特征。
Based on research for 955 maize particle images, the feasibilities of the maize image recognition and segmentation using RGB color pattern which were captured in the natural light and the filament lamp were discussed . By employing recognition software of the image, the geometric parameters of the maize image such as area, perimeter, diameter, roundness and elongation were extracted; Analysing on geometric characters parameters of the maize image, it was raised that using diameter and roundness to signify the maize geometric character of the maize.
出处
《鸡西大学学报(综合版)》
2006年第4期68-70,共3页
JOurnal of Jixi University:comprehensive Edition
关键词
玉米
图像处理
几何特征
maize
image processing
geometry property