摘要
针对黄花菜干菜分选设备缺乏、人工分选效率低、产品附加值不高等问题,研究基于机器视觉的黄花菜干菜分选方法。设计搭建了试验平台和图像采集系统,构建了干菜图像及尺寸信息基础数据库;基于数据转换、背景差分、全局阈值、轮廓提取等方法提取目标区域,采用外接矩形算法测量干菜相关尺寸。试验结果表明,本研究采用的方法能够准确测量黄花菜干菜长度,通过长宽比判断干菜品质的正确率较高,满足实际使用要求。
In order to solve the problems existing in sorting of dried daylily,such as the lack of equipment,low efficiency of manual sorting and low value-added products,this paper studied the sorting methods based on machine vision.The test platform and image acquisition system were designed and built,and the basic database of dried daylily image and size information was constructed.Then the target areas were extracted based on data conversion,background difference,global threshold and contour extraction.The circumscribed rectangle algorithm was used to measure the size of dried daylily.The experimental results showed that this method could accurately measure the length of daylily,and then judge the quality of dried daylily by length-width ratio,which could meet the actual use requirements.
作者
马聪
陈学东
Ma Cong;Chen Xuedong(Institute of Agricultural Economy and Information Technology,Ningxia Academy of Agriculture and Forestry Sciences,Yinchuan,Ningxia 750002)
出处
《宁夏农林科技》
2022年第4期61-65,共5页
Journal of Ningxia Agriculture and Forestry Science and Technology
基金
宁夏自然科学基金项目“基于机器视觉技术的黄花菜目标识别与定位方法研究”(2021AAC03257)。
关键词
黄花菜
全局阈值
机器视觉
外接矩形
Daylily
Global threshold
Machine vision
Circumscribed rectangle
