Underwater Object Detection in the Era of Artificial Intelligence: Current, Challenge, and Future

TL;DR

This survey comprehensively study AI-based UOD, categorising existing algorithms into traditional machine learning-based methods and deep learning-based methods, and summarising them by considering learning strategy, experimental dataset, utilised features or frameworks, and learning stage.

Abstract

Underwater object detection (UOD), aiming to identify and localise the objects in underwater images or videos, presents significant challenges due to the optical distortion, water turbidity, and changing illumination in underwater scenes. In recent years, artificial intelligence (AI) based methods, especially deep learning methods, have shown promising performance in UOD. To further facilitate future advancements, we comprehensively study AI-based UOD. In this survey, we first categorise existing algorithms into traditional machine learning-based methods and deep learning-based methods, and summarise them by considering learning strategy, experimental dataset, utilised features or frameworks, and learning stage. Next, we discuss the potential challenges and suggest possible solutions and new directions. We also perform both quantitative and qualitative evaluations of mainstream algorithms across multiple benchmark datasets by considering the diverse and biased experimental setups. Finally, we introduce two off-the-shelf detection analysis tools, Diagnosis and TIDE, which well-examine the effects of object characteristics and various types of errors on detectors. These tools help identify the strengths and weaknesses of detectors, providing insigts for further improvement. The source codes, trained models, utilised datasets, detection results, and detection analysis tools are public available at \url{https://github.com/LongChenCV/UODReview}, and will be regularly updated.

Figures (15)

• Figure 1: Challenges such as (a) degraded image quality and small objects, (b) noisy labels, and (c) class imbalance in underwater object detection datasets significantly impair the performance of deep detection models. For instance, (c) highlights that the deep detector RepPoints yang2019reppoints faces severe class imbalance issues on the DUO liu2021dataset dataset.
• Figure 2: The development of (a) AI in underwater object detection and (b) generic artificial intelligence.
• Figure 3: The frequency map of the keyword ‘underwater object detection' in Google Scholar from 2020 to 2024. The size of the keyword is proportional to the frequency of the word. The keywords ‘underwater object detection’, ‘deep learning’, and ‘neural network’ have drawn large research interests in the community.
• Figure 4: The road maps of (a) generic object detection and (b) underwater object detection reveal that the latter often leverages insights and techniques from the former to enhance detection performance in underwater envrionments.
• Figure 5: Comparisons of the sonar (top) and RGB (bottom) images. RGB images capture rich visual features but with limited perceptual range. Sonar images extend the perceptual range but are less intuitive and harder for humans to interpret.