远程控制采摘机器人手势识别研究——基于势场蚁群算法

发布时间：2018-09-12 20:12

【摘要】：采摘机器人在作业时遇到通过自主导航无法越过的障碍物时,或者在危险的地带无法进行人工采摘作业时,需要借助远程方式进行实时控制,使其成功越过障碍物,并在高危环境中有效地展开采摘作业。为了优化采摘机器人远程控制系统,提出了一种基于手势识别的远程控制方案,并引入了势场蚁群算法,提高了机器人的控制的准确性和高效性。在远程控制方案中,将基于视觉的手势识别与远程控制机械手相结合,通过深度相机采集手势图像并提取手势特征,转换为机械手舵机的控制命令,并通过无线网络发送至采摘机器人控制单元,实现视觉手势对机器人的远程控制。对采摘机器人进行了测试,通过测试发现:基于蚁群算法的手势识别系统可以有效地追踪得到不同的动态手势,且可以准确地识别手势所代表的意义,成功实现了机器人远程控制的手势识别。该方法不仅可以远程实现机器人避障功能,还可以将其应用在山谷、沼泽等危险地带进行采摘作业,实现其非凡的使用价值。
[Abstract]:When picking robots encounter obstacles that cannot be crossed by autonomous navigation, or when manual picking cannot be carried out in dangerous areas, real-time control by remote means is needed to enable them to successfully cross obstacles. And in the high-risk environment to effectively carry out picking operations. In order to optimize the remote control system of picking robot, a remote control scheme based on gesture recognition is proposed, and the potential field ant colony algorithm is introduced to improve the accuracy and efficiency of robot control. In the remote control scheme, the vision-based gesture recognition is combined with the remote control manipulator, and the gesture image is collected by the depth camera and the gesture feature is extracted, which is converted into the control command of the manipulator steering gear. The remote control of the robot by visual gesture is realized by sending the wireless network to the robot control unit. By testing the picking robot, it is found that the gesture recognition system based on ant colony algorithm can effectively track and obtain different dynamic gestures, and can accurately recognize the meaning of the gesture. The gesture recognition of robot remote control is realized successfully. This method can not only remotely realize the obstacle avoidance function of robot, but also can be used in picking in dangerous areas such as valleys and swamps, and realize its special use value.
【作者单位】：河南工业职业技术学院;
【基金】：河南省自然科学基金项目(2015ZCB115) 南阳市科技攻关项目(2012GG029)
【分类号】：TP242;TP18;S225

【参考文献】