Image water level intelligent identification method based on fixed viewport

Abstract

The invention discloses an image water level intelligent identification method based on a fixed viewport, for a picture shot by a fixed viewport camera, pixel points on the shot image have a one-to-one correspondence relationship with points on a corresponding external scenery. Position coordinates of each coordinate point of an external scene are acquired in advance, and the coordinates can be mapped to corresponding pixels on an image. When the external scenery changes regularly, such as rising or falling of the water surface of the water body, the edge line pixel of the change of the scenery in the identification area can be acquired by acquiring the fixed viewport image in real time and the intelligent algorithm, and the current water level value of the water body can be calculated bythe edge pixel. Sensor information such as images, water levels and rainfall can be automatically and regularly collected, file service and storage can be provided, parameter setting can be carried out on site or remotely, and automation of the whole process of water level carrier image acquisition, analysis and processing, water level intelligent recognition, information sending and the like canbe achieved.

Description

Technical field: [0001] The invention belongs to the technical field of hydrological monitoring, in particular to an image water level intelligent recognition method based on a fixed viewport. Background technique: [0002] Water level monitoring is one of the basic monitoring contents of rivers, lakes and seas. At present, traditional water level measurement sensors include: float water level gauge, bubble water level gauge, electronic water gauge, radar water level gauge, ultrasonic water level gauge, etc. No matter which type of water level measurement sensor has certain practicability and disadvantages. And were widely, good application. However, float-type water level platforms usually require a large investment and are difficult to construct, and are affected by unfavorable factors such as mainstream swing and sediment deposition; The extension cannot be too long, and the monitoring accuracy is easily affected by wind, waves, aquatic plants, floating objects, etc. On...

AI Technical Summary

Problems solved by technology

In this process, it is often accompanied by extremely complex machine learning calculations and noise filtering, and it also faces many technical difficulties. In the wild environment, water gauge reflections, water surface waves, mutual occlusion between water gauges, and water gauge Problems such as short water surface line segments on the signboard lead to low water level intelligent recognition and large water level errors. These problems have become important bottlenecks restricting the application of water level intelligent recognition technology

Images