Detection of the Fish Tail Frequency and Its Change Rules Based on Crucial Characteristics

Abstract

The tail frequency (the frequency of a fish's tail wagging) is an important parameter to describe the degree of the change of fish and evaluate the change of water quality. A tail frequency model that based on the crucial characteristic points of fish is proposed for detecting the tail frequency of the fish in real-time. Firstly, the target image is detected through image processing algorithm, and the binary image of the fish target area is obtained; then, according to the physical and tail characteristics of fish, the coordinate of centroid, tail point, head and other key points are obtained. The geometric model of the fish tail angle is established. Finally, the change rule of the tail frequency of the fish is obtained by the analysis of the fish tail angle in the continuous video frame. The experiments show that the fish tail characteristics in the single image can be effectively obtained and the fish tail frequency in pH = 7 (normal water quality), pH = 5 (abnormal water quality), pH = 3 (water quality deterioration) are consistent with the observed change rule of the tail frequency. The fish tail frequency characteristics can be effectively identified by the proposed model, and the theoretical basis and experimental method are provided for further studying about biological water quality.