自主行走葡萄收获机器人视觉导航信息处理技术研究

针对农副产品如何能够高效、高质量的收获这一大难题,本文对自主行走葡萄收获机器人视觉导航信息处理技术进行了深入的研究。首先,本文针对自主行走农业机器人的自主识别、收获农作物的问题,以葡萄收获机器人为模板,研究了基于视觉的导航信息处理技术。其次,分析了自主行走葡萄收获机器人的研究现状,因其对于重复性工作完成具有较高的质量和速度,相较于人工更可以保障在农产品收获阶段的品控。再次,给出了农业机器人控制器软件架构,分析了其主要特点。最后,提出了通过基于视觉SLAM的方法完成农产品的收获操作的具体方法。通过实验验证,与传统的激光导航算法相比,视觉导航技术具有更高的定位导航精度,定位精度可以2~3cm。证明了自主行走葡萄收获机器人视觉导航信息处理技术研究的有效性。

Research on visual navigation information processing technology of autonomous walking grape picking robot

In view of how the agricultural and sideline products can be charged with high efficiency and high quality,this paper has carried out in-depth research on the robotic visual navigation information processing technology for autonomous walking grapes.Firstly,this paper studies the self-identification of autonomous walking agricultural robots and collects crops,and uses the grapecollecting robot as a template to study the navigation-based navigation information processing technology.Secondly,the research status of the autonomous walking grape collecting robot is analyzed.Because it has higher quality and speed for the repetitive work,it can guarantee the quality control in the agricultural product collection stage compared with the artificial.Thirdly,the software architecture of the agricultural robot controller is given and its main features are analyzed.Finally,a specific method for completing the collection operation of agricultural products by the method based on visual SLAM is proposed.Through experimental verification,compared with the traditional laser navigation algorithm,the visual navigation technology has higher positioning and navigation accuracy,and the positioning accuracy can be 3~4cm.It proves the effectiveness of the research on robotic visual navigation information processing technology for autonomous walking grapes.