Abstract: In order to analyze the water quality rapidly and accurately,a new method of water quality anomaly detection is proposed.The method combines technique of visual perception with negative selection(NS) theory in artificial immune system.First,biological monitoring technique driven by vision is used to acquire water quality index.Second,algorithm of variable V-detector of NS theory is used to evaluate the water-quality index.Due to the sensition and comprehension of visual perception and self-organization,self-learning,self-identification of artificial immune system,the method can detect water quality rapidly and accurately.The experimental result shows that this method is an effective water quality assessment method.