摘要
传统番茄苗选择中,往往是操作员根据形态学特征,基于视觉评分标准来识别幼苗,这是一种主观的、容易发生人为错误的方法。针对这一问题,文章基于机器视觉技术设计了一种全自动移栽番茄苗分级分选算法,选择UXGA(Ultra eXtended Graphics Array)极速扩展图像阵列相机、蓝色背光和滤光装置获取番茄幼苗图像信息,从图像信息中确定幼苗的弯曲度、叶片节点和茎粗,然后对其进行分级和排序,进而将这些幼苗分级为“可接受”和“剔除”。结果表明,根据番茄苗茎的粗细,将幼苗分级,分为大、中、小三类,具有良好的效果。该算法不仅适用于番茄幼苗,还可以应用于茄子、辣椒和许多其他幼苗,分选成功率为97%。
Traditionally operators use a set of visual scoring criteria to identify seedlings based on morphological characteristics, and this is a subjective methoed prone to human error prone. To solve this problem, an automatic tomato seedling classification algorithm based on machine vision technology is designed by selecting UXGA camera, blue backlight and filter for tomato seedling image information. The algorithm is developed to determine the curvature, leaf node and stem diameter of seedlings from the image information, and then to classify and sort them. The seedlings have been graded as “acceptable” and “culled”. The results show that according to the size of the stem, the seedlings are classified into three types: large, medium and small. The algorithm is applied not only for tomato seedlings but also for eggplant, pepper and many other seedlings, with a success rate of 97%.
作者
张海歆
ZHANG Haixin(College of Artificial Intelligence, Yango University, Fuzhou 350015, China)
出处
《安阳师范学院学报》
2021年第2期23-27,共5页
Journal of Anyang Normal University
关键词
番茄幼苗分类
嫁接机器人
机器视觉
排序算法
classification tomato seedling
grafting robot
machine vision
sorting algorithm
