A Review on Tracking and Detecting Fish from Videos

Abstract

Non-extractive fish abundance estimation with the aid of visual analysis has drawn increasing attention. Unstable illumination, ubiquitous noise and low frame rate video capturing in the underwater environment, however, make conventional tracking methods unreliable. In this paper, we present a multiple fish tracking system for low-contrast and low-frame-rate stereo videos with the use of a trawl-based underwater camera system. An automatic fish segmentation algorithm overcomes the low-contrast issues by adopting a histogram back projection approach on double local-threshold images to ensure an accurate segmentation on the fish shape boundaries. The problem of non-uniform illumination over the video frame by focusing only on the vicinity of each target. The Slowly moving objects detection are present in the scene such problems. A New algorithm for detection and tracking will be implemented in order to investigate improved efficiency. Furthermore, the algorithms developed to perform the video analysis, (such as pre-processing, detection, tracking and counting) could be integrated into a more generic architecture so that the best algorithm for each step will be selected.