A real-time snowfall intensity estimation method based on video analysis

Abstract

The present invention discloses a real-time snowfall intensity estimation method based on video analysis, which is a real-time snowfall intensity analysis method. The real-time snowfall intensity estimation method includes five parts: region division, single-region snow particle detection, snow particle detection validity judgment, global snow particle estimation and real-time snowfall intensity estimation. Experiments show that by this method, distribution of falling snow particles can be effectively identified and calculation is conducted in real time. It is convenient to combine the method with existing video monitoring systems, therefore achieving high practical values.

Description

technical field [0001] The invention relates to the technical field of meteorological observation, and relates to a method for estimating real-time snowfall intensity based on video analysis. Background technique [0002] With the rapid development of high-speed railways, the speed of trains is getting faster and faster. Under extreme conditions, the safety of trains is also more important. Snowfall is the weather condition that has the greatest impact on the operation of railway trains. Snow accumulation will reduce the wheel-rail friction force and increase the braking distance, which will bring hidden dangers to driving safety. Therefore, real-time measurement of snow depth and real-time feedback of snow depth data to the control center provide an important basis for train scheduling. [0003] The traditional method is to use a snow measuring stick or a graduated snow measuring rod to be inserted into the snow to the ground surface to measure the snow depth on the ground...

AI Technical Summary

Problems solved by technology

[0003] The traditional method is to use a snow measuring stick or a graduated snow measuring rod to be inserted into the snow to the ground surface to measure the snow depth on the ground. Because there is usually an ice layer under the snow, it is difficult to measure accurately with the snow measuring stick, and at the same time, the snow caused by wind blowing Snow is redistributed, so representative measurement points and multiple measurements need to be selected, this method is time-consuming and laborious, has large random errors, and is difficult to monitor in real time

[0004] At present, the existing snow depth measurement methods include radioisotope method, which has high accuracy but will endanger the health of surveyors; the measurement method based on image processing is greatly affected by weather and light, and its robustness is poor; the ultrasonic sensor snow depth measurement method, its It is greatly affected by temperature and wind speed, and the snow depth measurement method based on the phase method laser ranging principle has strong stability but is affected by the unevenness of the snow surface

Images