基于机器视觉的鲜食葡萄采摘点和抓取点定位方法  

Table Grape Picking Point and Grasping PointPositioning Method Based on Machine Vision

在线阅读下载全文

作  者:马智杰 李长勇[1] Ma Zhijie;Li Changyong(School of Mechanical Engineering,Xinjiang University,Urumqi 830047,China)

机构地区:[1]新疆大学机械工程学院,乌鲁木齐830047

出  处:《农机化研究》2025年第7期191-197,共7页Journal of Agricultural Mechanization Research

基  金:新疆维吾尔自治区自然科学基金项目(2021D01C052)。

摘  要:在自然环境下准确快速识别葡萄并对采摘点和抓取点进行定位,对葡萄采摘机器人的收获过程至关重要。为此,针对葡萄在自然环境下难以准确地对葡萄的采摘点和抓取点进行定位的问题,提出了一种基于机器视觉的鲜食葡萄采摘点和抓取点定位方法。该方法使用YOLOv5模型来识别和分割出葡萄和果梗,然后利用求多边形的最小外接矩形的中心点来确定葡萄采摘点和抓取点的二维坐标;同时,结合结构光相机对采摘点和抓取点的三维坐标进行确定,在远距离粗定位葡萄,在近距离精确定位采摘点和抓取点。通过对青葡萄、紫红葡萄和紫黑葡萄3种不同颜色的葡萄进行验证,结果表明:葡萄和果梗的识别平均精确度分别为92.8%和89.8%,分割平均精确度分别为92.6%和92.7%,识别和分割后对采摘点和抓取点定位的成功率分别为95.9%和98.6%。Robotic grape harvesting depends on the precise and quick identification of grapes in their natural environment as well as the placement of picking and grasping points.Accurately locating the grapes′picking and gripping locations in the natural environment is challenging.For this,proposed a picking and grabbing point positioning approach for table grapes based on machine vision.The grape picking point and the gripping point were located using the center point of the smallest circumscribed rectangle of the polygon after the YOLOv5 model had been used to identify and segment the grapes and fruit stalks.The grapes were roughly positioned at a distance,and the picking point and gripping point were precisely located at a close distance,all while the three-dimensional coordinates of the picking point and the gripping point were calculated by merging the structured light camera.The average split rates of grapes and fruit stems were 92.6%and 92.7%,respectively,when the three various hues of grapes green grapes,purple red grapes,and purple black grapes were verified.The average recognition rates of grapes and fruit stems were 92.8%and 89.8%,respectively.After recognition and segmentation,the percentages of success in determining picking points and gripping point positioning were 95.9%and 98.6%,correspondingly.

关 键 词:采摘机器人 果梗识别 定位 YOLOv5 鲜食葡萄 

分 类 号:S225[农业科学—农业机械化工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象