Front-end water level detection system based on combination of online learning and offline learning

Abstract

The invention discloses a front-end water level detection system based on the combination of online learning and offline learning. Thefront-end water level detection system comprises an image acquisition module, an image transmission module and a system calculation module; the output end of the image acquisition module is connected with the image transmission module, and the output end of the image transmission module is connected with the system calculation module; the system calculation module comprises an off-line learning unit, an on-line learning unit and a data transmission unit, the off-line learning unit is connected with the on-line learning unit through the data transmission unit, and the image acquisition module further comprises a video acquisition unit which is used for carrying out image transmission when the water volume is increased too fast. According to the system, the detection process is simple and convenient, online learning and offline learning are combined, the environment generalization ability is improved, when the rainfall change is large, the analysis interval between video frames is automatically reduced, and the video clip at the specified time is generated, so that the real-time performance of water surface early warning is improved, and the key moment described by the video clip is further configured; and the system is suitable for popularization and use.

Description

technical field [0001] The invention relates to the technical field of water area detection, in particular to a front-end water level detection system based on the combination of online learning and offline learning. Background technique [0002] The water level refers to the elevation of the free water surface relative to a certain base surface, and the distance from the water surface to the bottom of the river is called the water depth. The base used to calculate the water level is a certain characteristic sea level elevation as the zero level base, called the absolute base, commonly used is the Yellow Sea base; you can also use the elevation of a specific point as a reference to calculate the zero point of the water level, called the station base noodle. The water level is the most intuitive factor to reflect the water regime of a water body, and its change is mainly caused by the increase or decrease of the water volume. In order to prevent or reduce the occurrence of ...

AI Technical Summary

Problems solved by technology

[0003] Water level detection is generally divided into online learning monitoring and offline learning detection. The advantage of online sensors is that they do not require offline labeling data and offline training, and the environment generalization ability is strong. The disadvantage is that the recognition accuracy is relatively low, and the offline learning scheme A large amount of data needs to be collected and labeled in advance, offline learning is also required, and the generalization ability of the environment is lower than that of online learning

Images