Research on Autonomous Grasping Control of Underwater Manipulator Based on Visual Servo

To address the problem associated with the low efficiency and intelligence level of underwater master-slave operation method, a grasping control method of underwater manipulator, on the basis of monocular vision servo, is proposed to improve the autonomous capability of underwater operation. Typical underwater objects can be grasped accurately and quickly. Firstly, ArUco marker is used to obtain accurate position information at a high rate of speed, attitude information is also calculated based on the shape and color of target in key frames from video stream at the same time. Grasping method is adopted automatically according to specific position and attitude of the target. Experiments are carried out with the 7-function underwater hydraulic manipulator developed by Shenyang Institute of Automation, Chinese Academy of Sciences. The manipulator is programmed to grasp the object automatically based on visual servo and place it to the appointed position. Both land and underwater experiments validated that the adopted control method of manipulator in this paper meets the actual needs of underwater operation.