Automated Quantification of Brittle Stars in Seabed Imagery Using Computer Vision Techniques.

Kazimieras Buškus,Evaldas Vaičiukynas,Antanas Verikas,Andrius Šiaulys,Aleksej Šaškov,Saulė Medelytė

doi:10.3390/s21227598

Abstract

Underwater video surveys play a significant role in marine benthic research. Usually, surveys are filmed in transects, which are stitched into 2D mosaic maps for further analysis. Due to the massive amount of video data and time-consuming analysis, the need for automatic image segmentation and quantitative evaluation arises. This paper investigates such techniques on annotated mosaic maps containing hundreds of instances of brittle stars. By harnessing a deep convolutional neural network with pre-trained weights and post-processing results with a common blob detection technique, we investigate the effectiveness and potential of such segment-and-count approach by assessing the segmentation and counting success. Discs could be recommended instead of full shape masks for brittle stars due to faster annotation among marker variants tested. Underwater image enhancement techniques could not improve segmentation results noticeably, but some might be useful for augmentation purposes.

Highlights

Underwater studies are critical from various aspects, such as economic, ecological, and scientific
In case the same mosaic is used both for training and testing, learning is performed on one-half of the mosaic by splitting it horizontally at the middle and training on the top while testing on the bottom halves
When training on mosaic-1, better results are achieved on the mosaic-2, whereas the same can not be said when training on mosaic-2

Summary

Introduction

Underwater studies are critical from various aspects, such as economic (off-shore wind farms and oil extraction platforms construction), ecological (biodiversity monitoring and impact assessment), and scientific (geology, archaeology, biology studies). One of the widely used seabed habitat mapping methods in the continental shelf and deep seas is underwater imagery [3,4]. Technological progress from hand-held cameras to remotely operated vehicles (ROV) and autonomous underwater vehicles (AUV) increases video material amounts and quality. This method’s main advantage is its simplicity, enabling the rapid collection of large amounts of data, and, cost-effectiveness. Only a small part of the information available in underwater imagery archives is being extracted due to labour-intensive and time-consuming analysis procedures, the need for automatic image analysis arises

Methods

Results

Discussion

Conclusion