Automated Image Processing for Fisheries Applications
National Oceanic and Atmospheric Administration (NOAA)
07/01/2016 - 06/30/2017
Automated image processing is a rapidly developing field in many areas of science and industry, especially the development of image-based analysis systems to identify, count, and measure fishes and other organisms, as well as to facilitate habitat classification. An automatic image processing system for fishery applications should include fish segmentation, length measurement, fish counting and tracking, fish species classification, and finally behavior understanding of fish schools. To extend out previous successful collaboration with AFSC on automated fish tracking, length measurement and species identification via the techniques of image processing and computer vision, in this proposal, we further propose to expand these efforts to more general challenges for automation, including multiple-target tracking and target shape segmentation from complex scenes captured by moving cameras.