Electronic monitoring has been shown to be an effective tool to meet a variety of fisheries monitoring objectives in compliance-based programs. However, these systems have not been effective in delivering individual fish data similar to information collected by an observer.
Development of new camera-based systems, methods, and tools is critical for collecting scientific data to inform management. A camera system being developed at the Alaska Fisheries Science Center greatly improves the functionality and addresses many of the limitations of electronic monitoring systems. This system provides the ability to automatically collect length measurements in addition to monitoring for compliance. System capacity to identify and automatically capture high quality (HD) stereo images of catch events, for efficient identification of fish to species or species group, is being actively developed. Because only images of individual catch events are stored and reviewed, post-processing and storage costs are reduced, facilitating data transfer and data management. Event-based image capture will also improve real time reporting, reducing lag times associated with current monitoring and post processing methods. Images will be linked to GPS information, allowing precise location of all species-specific catch. Spatial image tagging will enable mapping of high bycatch rate areas, potentially improving future management strategies to lower bycatch. Lowering costs associated with the collection, transfer, storage, and analysis of event-based image data proposed here will allow greater coverage rates to a wider range of vessel types and sizes where it is impractical to place an observer. By leveraging the latest developments in computer vision, cost-effective and timely extraction of scientific data from images will provide greater certainty for resource management and support sustainable fishing practices.
