Precipitation weather identification system based on intelligent identification technology

Abstract

The invention relates to the technical field of meteorological observation, in particular to a rainfall weather identification system based on an intelligent identification technology, which comprises an acquisition terminal, a preprocessing terminal, a training server terminal, a monitoring server terminal and a shooting terminal, the acquisition terminal, the preprocessing terminal, the training server terminal, the monitoring server terminal and the shooting terminal are connected through a network and are used for data transmission; according to the invention, the image or video of the weather is acquired through the acquisition terminal, then the image or video of the weather is labeled with the weather category label, and the image or video is preprocessed, so that the preprocessed image or video trains the floating-point model of the training unit; therefore, the image or video detection result of the trained model is more accurate, and the accuracy of the detection result is improved.

Description

technical field [0001] The invention relates to the technical field of meteorological observation, in particular to a precipitation-type weather recognition system based on intelligent recognition technology. Background technique [0002] In weather forecasting, the weather forecasting organization displays the data information of various weather elements to the user, informs the user of the change trend of the weather element, and helps the user understand the weather; in order to improve the efficiency and accuracy of the weather element data, the ground automatic meteorological observation station is the main The weather element data acquisition equipment collects various weather element data for weather forecast analysis, so as to predict the precipitation time and precipitation in the region; the method of ground automatic meteorological observation station to collect various weather element data includes physical sensor collection And image collection, the UAV is equip...

AI Technical Summary

Problems solved by technology

[0003] Existing methods for weather recognition include physical sensor recognition and image monitoring and recognition; physical sensors are low in cost, but can only provide a single type of data, and have lower requirements for outdoor environments. Compared with image monitoring and recognition, physical sensors can only provide a single Various types of data, weather information cannot be comprehensively analyzed, so the monitoring recognition rate is low, frequent maintenance and loss are more; the existing image information monitoring has high requirements on the outdoor (temperature, light) environment, and the ability to extract image classification features is weak , the learnability is not strong, so the detection of the weather is not accurate; and the carrier of the collected image information is easily disturbed by the external environment during work. When working in the outside world, dust will cause lens failure and damage, which will affect the collection of image information, resulting in a decline in the accuracy of weather recognition, which in turn will cause the limitations of the solution.

Images