Robot visual servo control system and method combining two arms with hand-eye camera

Abstract

The invention discloses a robot visual servo control system and method combining double mechanical arms with a hand-eye camera. The system comprises a robot trunk, double mechanical arms, a dexterous hand, a hand-eye camera, a camera light source and a robot controller. The two mechanical arms are installed on the robot body, the hand-eye cameras are installed on flanges at the tail ends of the mechanical arms, and the dexterous hands are installed at the tail ends of the mechanical arms. Diffuse reflection cooperative target two-dimensional codes are pasted on three side surfaces of the hand-eye camera. The robot controller is installed on the robot body and can control the robot to move. The robot controller is connected with the mechanical arm and the dexterous hand through an electronic circuit and controls the mechanical arm and the dexterous hand to move. The robot controller is connected with the two hand-eye cameras to obtain images. The robot controller is connected with the two camera light sources through an electronic circuit, and it is guaranteed that the cameras have the best illumination condition. The system covers far, near and ultra-near working distances, realizes high-precision non-cooperative target-oriented grabbing, saves the research and development cost and shortens the research and development period.

Description

technical field [0001] The invention belongs to the field of robot control, and relates to a robot visual servo control system and method combining dual mechanical arms and a hand-eye camera. Background technique [0002] The mobile dual-arm dexterous work robot needs high stability and accuracy when approaching and grasping non-cooperative targets. When approaching the target, it needs to rely on the sensor with a longer distance, and when grasping the target, it needs to rely on the sensor with a closer distance. [0003] At present, the existing robot's approach and grasping technology for non-cooperative targets uses traditional position sensors and force sensors, which are affected by the installation gap of the manipulator arm and have a certain degree of flexibility, and the accuracy is low. The hand-eye separation technology is to install the vision equipment and the robotic arm independently on the robot. In this case, the fixed camera makes it difficult to cover al...

AI Technical Summary

Problems solved by technology

[0003] At present, the existing robot approaching and grasping technology for non-cooperative targets uses traditional position sensors and force sensors, which are affected by the installation gap and flexibility of the robotic arm, and the accuracy is low.

The hand-eye separation technology is to install the visual equipment and the robotic arm independently on the robot. In this case, the fixed camera makes it difficult to cover all working distances, far and near, and the flexibility is not enough; the hand-eye integration technology installs the visual equipment on the robotic arm. At the end, in this case, the dexterous hand at the end of the robotic arm may block the camera when performing fine grasping, resulting in a blind spot of the camera. When performing ultra-close-range operations, the camera is likely to lose sight of the target

Using a zoom camera to observe the target, because the focal length value of the zoom camera changes greatly, has a great impact on the stability of the measurement camera, which will reduce the measurement accuracy; using a binocular camera to provide measurement data will increase R&D costs and prolong the R&D cycle

Images