Artificial intelligence water level monitoring system based on SVM water gauge online binary classifier

Abstract

The invention discloses an artificial intelligence water level monitoring system based on an SVM (Support Vector Machine) water gauge online binary classifier, a working method of the artificial intelligence water level monitoring system comprises training of the online binary classifier and water level measurement, and the training of the online binary classifier comprises the following steps: Step 1: configuring a reference image; Step 2, extracting image features; Step 3, classifying pixel points to obtain a classification graph; Step 4, drawing up a water level line; and Step 5, obtaining a binary classifier. The system has the beneficial effects that the water level value is monitored according to the SVM-based online binary classifier, and the artificial intelligence degree is high; accuracy is high and calibration is rapid; the real-time degree is high, and information updating is fast; the monitoring system can monitor various surface water in reservoirs, rivers, lakes, urban water areas and the like in various states, and is wide in monitoring application range.

Description

technical field [0001] The invention relates to the field of water level monitoring, in particular to an artificial intelligence water level monitoring system based on an SVM water scale online binary classifier. Background technique [0002] It is difficult to identify the water gauge in snowy and foggy weather. In most cases, the water gauge image is invisible and cannot be recognized. However, this weather generally occurs in the dry season, so it is not considered for the time being. In rainy days (drizzle, light rain, moderate rain, heavy rain, heavy rain), the visibility of the actual image water scale will prevail, and there will be varying degrees of blur and occlusion. Similarly, using a similar E-type water scale will enhance the anti-interference to a certain extent. Draft gauge specifications, interference and reliability. According to the image data provided by the monitoring points, non-standard water gauges have different specifications, and there are varying...

AI Technical Summary

Problems solved by technology

According to the image data provided by the monitoring points, non-standard water gauges have different specifications, and there are varying degrees of degradation. In addition, the site environment is complex, and the impact of lighting, shadows, water waves, water stains, stains, garbage, etc. The reliability of using the same recognition algorithm for different monitoring points is not high, and the feasibility of customizing algorithms for different monitoring points is not high.

However, there is a large error in the way the human eye recognizes the water gauge, and the efficiency is too low

Images