基于计算机视觉的萝卜幼苗自动识别技术

Application of Computer Vision in Automatically Discriminating Radish Seedlings

为实现萝卜幼苗间苗作业的自动化 ,提出并探讨了适于萝卜幼苗自动间苗的播种方式。通过对萝卜幼苗彩色图像的分析 ,提出了由突出绿色方法和亮度组合的灰度化方法、最大方差自动取阈法及快速区域标记等构成的自动识别算法。对从田地中拍摄的 2 5 0幅萝卜幼苗图像的识别处理结果表明 。

The research is to develop a robotic system for thinning out radish seedlings as its overall objective. This paper proposed and discussed a new planting way suitable to automatic thinning and an automatic discrimination algorithm of radish seedlings based on computer vision. The algorithm includes combination of the ExG (Extra Green) and brightness, automatically thresholding, and fast region labeling and so on. Experiments performed on radish seedling images taken from the farm showed that radish seedlings were fully discriminated from background in all of images of 250 sets of seedlings.