排球运动规划在采摘机器人识别定位中的应用

Application of the Volleyball Motion Planning in the Recognition and Location of Picking Robot

为对采摘机器人的识别与定位功能进行优化,将排球机器人的运动规划原理与采摘机器人的控制要求相结合进行应用探讨。通过搭建采摘机器人对果实的识别定位理论模型,运用核心图像识别与处理算法,硬件配置动作执行协调及软件系统后台指令控制,实现多功能性传感装置信息数据的合理性采集与传输,达到实时定位目标。进行了采摘机器人的识别与定位试验,结果表明:在排球机器人运动规划与控制机理下,通过目标与定位图像的有效抓取,采摘定位时间可控制在0.6s左右,综合定位准确率保持在93.8%以上,最高定位准确率可达95.7%,满足采摘机器人作业需求,验证了设计理念的可行性,可为类似农业设备定位开发提供思路。

In order to optimize the recognition and positioning function of the picking robot in depth,the motion planning principle of the volleyball robot and the control requirements of the picking robot were combined into discuss.By building the theoretical model of fruit recognition and positioning for picking robot,using the core image recognition and processing algorithm,the coordination of hardware configuration and action execution and the background instruction control of software system,the rationality of collecting and transmitting information data of multi-functional sensor device was realized,then the target of the accurate and real-time positioning was achieved,and the recognition and positioning experiments of the picking robot were carried out.Experiments showed that under the mechanism of motion planning and control of volleyball robot,the whole process of recognition and positioning processing of picking robot was understood through effective grabbing of target and positioning image.On the other hand,picking and positioning time could be controlled at about 0.6 seconds,and the comprehensive positioning accuracy could be maintained at over 93.8%,the highest positioning accuracy could reach 95.7%,which could meet the picking operation requirements of the picking robot and verify the feasibility of the design concept,also provide some ideas for the development and improvement of similar agricultural equipment positioning.