基于深度双目视觉处理的智能采摘机器人设计OA

Aiming at the problem that picking robot's location of fruits and vegetables was not accurate enough and it could not avoid obstacles accurately,which leaded to low picking efficiency,the intelligent picking robot based on deep binocular vision processing was designed.The intelligent picking robot was constituted of PLC controller,vision system,mobile platform,navigation system,mechanical arm,communication system and power supply.To carry out optimal path planning and obstacle avoidance for the mechanical arm of the picking robot,the binocular vision system was used to ac-curately locate fruits and vegetables after the pre-processing of the collected images,and then the method of Huff trans-form linear detection was used to design and select the optimal path,and finally determine the optimal picking path.The picking robot movement trajectory accuracy test and picking test were done,the test results show that the picking robot for fruit and vegetable picking success rate is high,and it could meet the requirement of fruit farmers for picking robot.