Three-dimensional visual mapping of underwater ship hull surface using image stitching geometry

Abstract

This paper presents a method for three-dimensional (3D) visual mapping of underwater ship hulls using monocular vision. Visual inspection of underwater ship hulls using unmanned underwater vehicles needs to be conducted at very close range to the target surface because of the limited visibility in underwater environments. These environmental challenges can result in ineffective photometric and geometric information in hull surface images. Therefore, the performance of conventional 3D reconstruction techniques based on monocular vision is often unsatisfactory. To address these challenges and generate a precise 3D global map of the surveyed hull surface, the proposed visual mapping method adopts a two-step approach comprising global registration in two-dimensional space and 3D reconstruction using the information of the stitching geometry drawn by the registered images. Experimental results are shown to demonstrate the feasibility of the proposed method using image datasets obtained in field experiments with two different full-scale ships in real sea environments.