作 者：
梁秀梅;刘涛;孙成明;郭斗斗;卢岩;陈雯;孙洋洋
单 位：
扬州大学农学院江苏省作物遗传生理国家重点实验室培育点
关键词：
作物籽粒;图像处理;计数;分水岭算法;综合算法
摘 要：
图像识别技术是计算机视觉技术的重要内容之一。为了解决传统作物籽粒计数方法的不足,以水稻籽粒为试验对象,通过外部设备获取高质量的籽粒图像,利用计算机图像处理技术设计配套的软件,并改进相应的图像分割算法,可以快速准确识别出水稻籽粒数量,准确率可达98%以上。结果为计算机视觉识别技术在农业上的进一步应用奠定了理论与实践基础。
译 名：
The Recognition of Crop Grains Based on the Computer Image Technology
作 者：
LIANG Xiu-mei et al(Key Laboratory of Crop Genetics and Physiology of Jiangsu Province,College of Agriculture,Yangzhou University, Yangzhou,Jiangsu 225009)
关键词：
Crop seed;Image processing;Counting;Watershed algorithm;Integrated algorithm
摘 要：
Image recognition technology is one of the important contents of computer vision technologies.In order to solve the shortage of the traditional crop seed counting methods,the rice grains were selected as experiment object in this research,and the high quality grain images were obtained through an external device.We designed the matching recognition software by using computer image processing technology,and improved the corresponding image segmentation algorithm.So the rice grain number could be identified quickly and accurately,and the accuracy could reach above 98%.The results established the theory and practice foundation for the further application of computer vision recognition technology in agriculture.
