Using Underwater Video to Monitor Fish Around Ocean and River Energy Devices

July 2, 2018, 17:00 - 18:30 UTC (9:00 am PDT/12:00 pm EDT/5:00 pm BST)

Summary
This webinar will discuss guidelines for using underwater video effectively and present a new software tool developed specifically for underwater video analysis. The software, called EyeSea, is being developed to help manage the video analysis workflow and to automate the detection of fish. Camera and recording equipment will be discussed and guidelines will be presented for recording video that is optimized for automated processing.

Presenters

Shari Matzner, Senior Engineer at Pacific Northwest National Laboratory
Dr. Matzner has been developing technology for marine monitoring, particularly around wind and marine renewable energy devices, for the last seven years.  She has developed algorithms for the detection and classification of signals in underwater video, thermal video, passive acoustics, active acoustics, and satellite imagery.  Her work includes the ThermalTracker software that automatically detects birds and bats in thermal video and the Nekton Interaction Monitoring System, a real-time sonar processing system for detecting and tracking fish and marine mammals.

Darren Odom, Chief Executive Officer at Boulder AI
With a background in product development, vision systems integration, and computer vision, Darren Odom has decades of experience in consulting and engineering. Before founding Boulder AI, Darren worked as an engineer team leader in vision systems, with projects in a diverse array of industries. Darren has experience in engineering management and consulting, and has worked in satellite communication, medical device R&D, and test systems development.

A video recording of this webinar is available here:

Comments

Submitted by Jonathan Whiting on Tue, 08/14/2018 - 05:35

Permalink

Question from Sean Eagan: Dr. Matzner, how does a human or computer decide if they same fish swam infront of the computer multiple times?

Submitted by Shari Matzner on Mon, 08/20/2018 - 23:08

Permalink

Sean, there is no way to tell if a fish that is in the camera's field of view is the same fish that was in the field of view at an earlier time. This is important to keep in mind when using camera-based study methods (acoustic methods have the same problem.) Depending on the study objectives and on the setting (river vs. open water), researchers can use the raw counts from video to infer species abundance or population density.