摘要
对果实图像进行实时、准确的目标提取是猕猴桃采摘机器人研究的关键技术。本文在对各种背景图像进行颜色特征分析的基础上,利用RGB颜色模型中的R-B颜色分量实现图像分割,并通过对不同的阈值方法进行研究对比,最终确定采用阈值为93的固定阈值方法进行图像分割,实验表明其分割效率达到82%以上。
For the kiwifruit harvesting robot, to extract fruit object from images quickly and exactly are the key technologies. Based on the color analysis from the image which kiwifruit in the nature background, utilized the R-B channel information in the color model of RGB to implement the segmentation of kiwifruit, and through the research and analysis of different threshold methods, finally adopt the result for the static threshold is 93 to separate the kiwifruit exactly from the background, with experiment indicate the segmentation effect is greater than 82 %.
出处
《微计算机信息》
2009年第18期294-295,277,共3页
Control & Automation
关键词
猕猴桃采摘
颜色模型
图像分割
目标提取
kiwifruit picking
color model
image separate
object extraction
