Object grabbing method and device based on 3D matching and computing equipment

Abstract

The invention discloses an object grabbing method and device based on 3D matching and computing equipment, and the method comprises the steps: obtaining a scene image and a point cloud corresponding to the scene image, inputting the scene image into a deep learning segmentation model, and carrying out the instance segmentation processing to obtain the segmentation result of each object in the scene image; according to the point cloud corresponding to the scene image and the segmentation result of each object, determining the point cloud corresponding to each object; for each object, matching the point cloud corresponding to the object with a preset template point cloud, and determining pose information of the object; and according to the point clouds corresponding to all the objects, determining the stacking relation between the objects, determining a target object from all the objects according to the stacking relation, converting the pose information of the target object into a robot coordinate system, and transmitting the converted pose information of the target object to the robot. According to the scheme, the pose information of each object is accurately determined, and the object grabbing accuracy is effectively improved.

Description

Technical field[0001]The present invention relates to the field of computer technology, and more particularly to a 3D matching object grab method, apparatus, and computing device.Background technique[0002]With the development of industrial intelligence, it is increasingly popular by robots in order to operate in artificial objects such as industrial parts, boxes, etc.). When the robot is operated, it is generally necessary to capture the object, move the object from one position and placed to another location, for example from the conveyor belt to move the object and placed on the tray or in the cater, and crawl from the tray The object is placed on the conveyor or other tray as required. However, in the prior art, the determination of the bites information of the object to be grasped is not accurate, which is easy to cause the robot to crawl the mistakes, such as the robot can't succeed in grabbing the object, or the object is dropped and even caught. Taken by the pressure body res...

AI Technical Summary

Problems solved by technology

However, in the prior art, the determination of the pose information of the object to be grasped is not accurate enough, which may easily cause the robot to make a grasping error. For example, the robot cannot successfully grasp the object, or the object falls after being grasped, or even What is taken is that the pressed object causes the object above the pressed object to fall, etc., which affects the realization of industrial automation

Images