Online Monitoring System of Fish Behavior

In order to overcome the defect of physic-chemical monitoring system and enhance the intelligence of biological monitoring technology, an online monitoring system of fish behavior is proposed in this paper. The main contributions of this system are as follows: 1) adaptive background updating algorithm (ABU); 2) automatic camshift tracking algorithm (ACT) with twice searching; 3) particle filter tracking (PFT) algorithm; 4) persistent turning walker (PTW) model; 5) artificial immune algorithm (AIA). This online system is used to monitor and analyse the fish behavior continuously, establish a normal behavior model and detect the anormal behavior. Experimental results show that the system is running stably and has achieved well effect in the simulation environment.