Robot dynamic grabbing method and system based on global and local visual semantics

Abstract

The invention discloses a robot dynamic grabbing method and system based on global and local visual semantics. The method comprises the following steps of identifying the object category and object size of a to-be-grabbed object, acquiring the position and posture of the to-be-grabbed object in a camera coordinate system, and determining the size of a gripper; and planning a static trajectory of amechanical arm through a global binocular camera and planning a the dynamic trajectory of the mechanical arm through a local binocular camera for grabbing the to-be-grabbed object, and achieving optimal grabbing. Through the method, the positioning precision of the position and posture of the grasped object is improved, the strength and size of the gripper are adaptively adjusted, dynamical planning and disturbance response are achieved, meanwhile, an object identifying algorithm is simplified, and the object identifying precision is improved.

Description

technical field [0001] The invention relates to the technical field of robot visual servoing, in particular to a robot dynamic grasping method and system based on global and local visual semantics. Background technique [0002] In the related technology, there are two ways to realize the grasping of objects, among them, the first static method: the camera is fixed on the rack as a global camera, taking pictures at fixed points, obtaining pictures, analyzing and processing them, and calculating the grasping points The position and posture of the robot are transmitted to the robot controller, and the robot arm executes the predefined grasping process. Similarly, the camera can be fixed at the end of the robotic arm as a local camera, and the robotic arm moves to a certain position to take fixed-point photos and execute the same predefined grabbing process. The second dynamic method: the camera is fixed at the end of the robotic arm. The visual servoing method is used to trac...

AI Technical Summary

Problems solved by technology

[0003] However, at present, in robot grasping, the data is usually obtained singlely, from the global camera fixed on the rack or from the local camera fixed at the end of the robot arm, and the multi-sensory interactive data is rarely processed.

Insufficiency of a single static grasping method and dynamic grasping method: the static grasping method cannot dynamically respond to the external environment, such as the falling of the grasped object, and the re-dynamic trajectory planning cannot be performed

[0005] However, the traditional visual recognition method requires manual design of a large number of operators, which is too inefficient and the recognition accuracy is not high enough.

Images