Intelligent handling robotic arm system based on 3D vision and deep learning, and using method

Abstract

The invention proposes an intelligent handling robotic arm system based on 3D vision and deep learning. The system comprises a vision detection module, a training learning module, a motion planning module and a control module, wherein the vision detection module collects images of objects and sends the images to the control module; the training learning module collects sample data and forms a database for the objects that the robotic arm needs to perform grasping actions; the motion planning module comprises a path planning part and a grasping motion planning part, and the path planning part realizes the path planning of the robotic arm and realizes the function of autonomous selection of paths and obstacle avoidance of the robotic arm; the grasping action planning part realizes the grasping function; and the control module processes the information transmitted by the vision detection module, the training learning module and the motion planning module and transmits corresponding commands to the vision detection module, training learning module and motion planning module, so that the robotic arm completes path movement and grabbing. The invention makes the work scene more diversified, the production and transportation more intelligent, and the application field of the robotic arm broadened.

Description

technical field [0001] The present invention relates to image processing, artificial intelligence and manipulator technology, and in particular to an intelligent handling manipulator system based on 3D vision and deep learning and its usage method. Background technique [0002] Patent Publication No. CN 108789414 A discloses a three-dimensional machine vision-based intelligent manipulator system and its control method, which uses a laser ranging module to detect objects. Through the projector, camera and camera, the stripes of the projector on the object are used to collect the size and position of the object. Then, according to information such as time and size, the grasping position of the robotic arm is automatically adjusted, so that intelligent grasping and processing of objects of different specifications can be realized. The use of projected fringe size inspection has high requirements for the light requirements of the working scene and the camera, and this method do...

AI Technical Summary

Problems solved by technology

This method is far from using artificial intelligence to achieve sorting efficiency, and it does not explain the robotic arm grasping and path planning

Images