Fundamental design of an autonomous underwater vehicle with camera-based object detection

Abstract

The autonomous underwater vehicle (AUV) has received considerable attention within the last decades. In this paper, few aspects were involved in the designing of an AUV including mechanical structures design, selection of materials, use of electronics and sensors and software algorithms. The algorithms of software were divided into three parts which are object detection, controlling and path planning. This paper aims to review the state-of-the-art techniques in underwater image processing in real-time. The algorithm applied for the object recognition in the study was You-Only-Look-Once (YOLO) by Joseph Redmon. This research paper collaborated with Fundamental Design of an Autonomous Underwater Vehicle with Sonar-based SLAM by Siddesh Govind. The details of the path planning by using SLAM and electricals overview were written in the collaborative paper.