Research Of Fish Detection And Tracking Based On Video

In recent years, it has become a hot issue to analyze animal behavior with computervision and image processing techniques. The fish behavior monitoring based on videosequences and the development of fish bio-robot has drawn lots attention. At present,getting the fish behavior data in a fast, accurate and stable way is a great challenge, andcan be used in many practical applications.Based on computer vision technology, the research in the paper is to detect and trackfish moving target. Focused on the acquisition of fish position and the detection of motiondirection, the CAMShift algorithm for fish trajectory tracking in video sequences isimproved. The major study is summarized as follows:Based on the study of the moving target detection algorithm, a segmentationalgorithm combining the background subtraction method and the OTSU algorithm is usedto get a good segmentation image of fish. Based on the result, we got the real-timelocation of the fish with the first moments.An automatic detection method of fish movement direction is studied in this paper.First, based on the obtained moving fish object, the fish contour and skeleton is extracted.Then, the angle between the skeleton and image axes is calculated. Finally, according tothe criteria of the direction, the direction of fish movement is obtained automatically.Based on the traditional CAMShift algorithm, an improved CAMShift algorithmcombined with motion segmentation is proposed in this paper. The algorithm is to trackthe moving fish target in video, and extract the trajectory of fish movement.The experiments showed that the proposed method in this paper can detect themoving fish target accurately and get the location and the direction of the fish movementautomatically. It presented the state of stable tracking, and accessed the trajectory.