Water level monitoring method based on dictionary learning

Abstract

The present invention discloses a water level monitoring method based on dictionary learning. The method comprises the following steps of: the step 1, fixing a measuring rod at a river area to be measured, and collecting images of the measuring rod and areas near the water surface in real time through a camera; the step 2, collecting a plurality of images as training images, dividing each image into C classes, employing a sliding window to collect training data for each class of image blocks, and marking corresponding class tags; the step 3, employing a dictionary learning method to perform training of the training data to obtain a dictionary D; and the step 4, employing the dictionary D to perform real-time classification of the images collected by the camera to obtain class tag vectors of the images so as to obtain a water level value. The water level monitoring method can be achieved only through installation of a traditional rod so as to overcome the problem that the horizontal plane position is difficult to accurately identify in a traditional water level measurement scheme, and has high stability and robustness.

Description

technical field [0001] The invention relates to the field of computer vision image processing, in particular to a water level monitoring method based on dictionary learning. Background technique [0002] Since water security and water resource issues seriously affect the sustainable development of the economy and society and the safety of people's lives and property, it is particularly important to monitor the water levels of rivers, lakes, rivers, reservoirs, etc. in real time. The current water level monitoring technology mainly obtains the water level value through manual real-time observation by establishing a video monitoring system in these areas. [0003] Most of the video monitoring methods only use the real-time browsing function of the video. Especially for the monitoring of flood disasters in the rainy season, 24-hour uninterrupted observation is still required manually, which consumes a lot of manpower and material resources, is inefficient, and the method of man...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the deficiencies in the prior art, the present invention provides a water level monitoring method based on dictionary learning, which solves the problem that the traditional water level monitoring method requires human eyes to observe the water level with low efficiency and difficulty in distinguishing

Images