Automatic coral reef fish identification and 3D measurement in the wild

TL;DR

This paper introduces a stereo image-based pipeline for automatic identification, 3D tracking, and measurement of fish populations in natural coral reef environments.

Contribution

It presents a novel method combining stereo imaging and automated analysis for real-time fish identification and 3D measurement in the wild.

Findings

Successful automatic fish identification in natural habitats

Effective 3D tracking and measurement of fish populations

Potential for real-time ecological monitoring

Abstract

In this paper we present a pipeline using stereo images in order to automatically identify, track in 3D fish, and measure fish population.