A Feature Learning and Object Recognition Framework for Underwater Fish Images

TL;DR

This paper introduces an unsupervised feature learning and error-resilient classification framework for underwater fish image recognition, addressing challenges like poor image quality and environmental variability to improve accuracy.

Contribution

It presents a fully unsupervised feature learning method combined with a hierarchical classifier and partial classification, advancing automation and robustness in underwater fish recognition.

Findings

Achieves high accuracy on public and self-collected datasets

Handles class imbalance and high uncertainty effectively

Introduces partial classification for ambiguous images

Abstract

Live fish recognition is one of the most crucial elements of fisheries survey applications where vast amount of data are rapidly acquired. Different from general scenarios, challenges to underwater image recognition are posted by poor image quality, uncontrolled objects and environment, as well as difficulty in acquiring representative samples. Also, most existing feature extraction techniques are hindered from automation due to involving human supervision. Toward this end, we propose an underwater fish recognition framework that consists of a fully unsupervised feature learning technique and an error-resilient classifier. Object parts are initialized based on saliency and relaxation labeling to match object parts correctly. A non-rigid part model is then learned based on fitness, separation and discrimination criteria. For the classifier, an unsupervised clustering approach generates a binary class hierarchy, where each node is a classifier. To exploit information from ambiguous images, the notion of partial classification is introduced to assign coarse labels by optimizing the "benefit" of indecision made by the classifier. Experiments show that the proposed framework achieves high accuracy on both public and self-collected underwater fish images with high uncertainty and class imbalance.