A water quality toxicity detection method based on fish activity analysis

Abstract

A water quality toxicity detection method based on fish activity analysis, comprising the following steps: 1) using crucian carp as a biological monitoring object to monitor it in real time; crucian carp group, the tracking video sequence of crucian carp is obtained; 3) the motion data analysis and detection of crucian carp includes the following process: 3.1), the difference between the movement speed of crucian carp, the contour area and the area mean value is used as the main feature data; 3.2), based on the simple shell The Yeesian classifier algorithm generates a mature detection model; 3.3), use new feature data to detect and judge whether the detection model is mature; 3.4), use the mature detection model to detect real-time water quality data online, and finally realize the online detection of water quality toxicity. It can realize online real-time detection, improve the sensitivity and continuity of water quality detection, reduce the detection cost, and can effectively detect a large number of unknown water quality toxicity conditions in real time.

Description

technical field [0001] The invention relates to the fields of biological water quality detection, HSV space domain transformation and extraction contour, naive Bayesian classification technology and the like, and proposes a biological water quality toxicity detection method. Background technique [0002] Water quality anomaly detection is based on the normal data index of the monitored water quality, and determines whether an abnormal water quality has occurred by checking the deviation degree between the current data index of water quality and the normal data index. Water quality detection can be simplified as a classification problem, the input is water samples, and the output is normal water quality or abnormal water quality. [0003] Most of the traditional water quality testing methods use physical and chemical testing combined with statistical model analysis to extract and test water quality data. The shortcomings of the traditional method are: 1. Most of the water qu...

AI Technical Summary

Problems solved by technology

The shortcomings of the traditional method are: 1. Most of the water quality data collection is manually collected regularly, resulting in a lag in collection and cannot cope with the detection of sudden water pollution; 2. Due to the use of physical and chemical detection, a series of analysis, Testing equipment, resulting in higher testing costs

Images