Research Challenges, Recent Advances and Benchmark Datasets in Deep-Learning-Based Underwater Marine Object Detection: A Review

Abstract

<p>Underwater marine object detection, as one of the most fundamental techniques in the community of marine science and engineering, has been shown to exhibit tremendous potential for exploring the ocean in recent years. It has been widely applied in monitoring of underwater ecosystems, exploration of natural resources, management of commercial fisheries, and so on. However, due to complexity of the underwater environment, characteristics of marine objects, and limitations imposed by exploration equipment, detection performance in terms of speed, accuracy, and robustness could be degraded dramatically when conventional approaches are used. Deep learning algorithms have been found to have significant impacts on a variety of engineering fields, including marine engineering. In this context, we offer a review of deep-learning-based underwater marine object detection techniques. To facilitate a thorough understanding of the subject, we organize research challenges of underwater object detection into three categories, namely image quality degradation, small object detection, and poor generalization. We review recent advances in underwater marine object detection and highlight advantages and disadvantages of existing solutions for each challenge. In addition, we examine the most extensively used benchmark datasets in detail and critically. Comparative studies with previous reviews, notably approaches that leverage artificial intelligence, and future trends on this hot topic are also presented.</p> <p><br></p> <p><strong>* This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.</strong></p>