Computer vision and machine learning algorithms for detecting marine life from acoustic backscatter time series

点击翻译按钮获取中文摘要

Our proposed research explores novel ways of detecting visual patterns from echosounder data usingcomputer vision and machine learning techniques. Switching from the traditional acoustic data processingparadigm to image analysis and visual learning-based algorithms may provide new insights into echosounderdata, and may enable our partner company to enrich their portfolio of automatic data processing services.Our industrial partner is ASL Environmental Sciences Inc (ASL), an employee-owned company based inVictoria, BC. ASL is a global leader in innovative solutions for environmental monitoring and has over 40years of experience in oceanographic, acoustic, remote sensing and ice research products and scientificconsulting services. Since 2000, ASL has been building acoustic backscatter echosounders. Until now, ASLhas provided very limited data processing software related to its echosounders. Adding new data products toASL's portfolio addresses an important need of its larger base of customers, which relates to their ability tointerpret and analyze higher amounts of complex data.This proposed research builds on encouraging results obtained during a current NSERC Engage grant that hasenabled us to establish a first collaboration with ASL. We will address the limitations identified during theEngage project and dig deeper into the problem of detecting marine life from echosounder data, using imageprocessing, deep learning and pattern recognition techniques.