A review on visual servoing for underwater vehicle manipulation systems automatic control and case study

Abstract

Visual servoing can greatly improve underwater robot manipulation accuracy and automation. Its main purpose is to control the end-effector pose of underwater vehicle manipulator systems (UVMS) relative to the target by using features extracted from image. However, underwater manipulation visual servoing suffers from the drawback of perception limitation of visibility range, obscure and low contrast image, inaccurate camera calibration and model parameters. This article has reviewed the state of major visual servoing method on UVMS including position based visual servoing (PBVS) and image based visual servoing (IBVS). On the problem of underwater manipulation and vision control, a novel hybrid strategy of visual servoing control based on uncalibrated visual servoing has been issued. Simulations and comparisons on PBVS, IBVS and proposed hybrid strategy have been made. The article is concluded with challenge analysis and future research trends discussions.