Fish motion detecting algorithms based on computer vision technologies

In order to improve the detecting accuracy of fish motion, based on the analysis and comparison of classical moving object detection algorithms, three improved detection methods suitable for moving fish detection were presented. Experimental results showed that the proposed fish motion detecting algorithms had advantage over traditional methods. Especially, moving average background and the temporal difference method was better than other algorithms in detecting time and detecting accuracy, it has great potential in fish motion detection.