A cloud detection method, system and electronic equipment of Fengyun satellite remote sensing image

Abstract

The present application relates to a cloud detection method, system and electronic equipment of Fengyun satellite remote sensing images. Including: step a: extracting the infrared 1 band and visible light band of Fengyun satellite remote sensing images, and using at least two threshold methods to obtain at least two binary threshold results of the infrared 1 band and visible light band respectively; step b: using integrated The threshold method integrates at least two binarized threshold results of the infrared 1 band and visible light band respectively to obtain the integrated threshold cloud detection result of the remote sensing image; step c: according to the integrated threshold cloud detection result and the official provided The cloud detection results jointly select a random forest training sample set; step d: construct a random forest detection model through the training sample set, and perform cloud detection of Fengyun satellite remote sensing images through the random forest detection model. The application effectively improves the misdetection and missed detection of clouds, and improves the utilization rate of Fengyun satellite remote sensing images.

Description

technical field [0001] The present application belongs to the technical field of remote sensing image cloud detection, and in particular relates to a cloud detection method, system and electronic equipment of Fengyun satellite remote sensing image cloud. Background technique [0002] With the continuous development of imaging technology, there are more and more channels for remote sensing image acquisition. In the process of wind and cloud remote sensing image acquisition, due to the influence of cloud and fog interference and other factors, the spectrum of the original object is distorted, which affects remote sensing products and image interpretation, and has a great impact on information extraction. Correctly separating cloudy pixels and cloudless pixels in remote sensing images has a significant impact on weather forecast, prevention of meteorological disasters, temperature inversion, rescue and monitoring of ecological environment. Therefore, in the process of using re...

AI Technical Summary

Problems solved by technology

The frequency combination threshold method mainly utilizes the low-frequency characteristics of clouds, and obtains image low-frequency data through wavelet analysis, Fourier transform and other methods for cloud detection. Detection efficiency

[0005] 2) Texture analysis method: Texture analysis method utilizes the difference between cloud and ground texture features, usually in units of block sub-images, combined with second-order moment, fractal dimension, gray level co-occurrence matrix and multiple bilateral filtering to calculate texture features. Class methods need to obtain reliable cloud feature intervals in advance to ensure the accuracy of classification, and the efficiency is low

In the process of implementing the comprehensive intelligence method, a large number of training samples need to be obtained, and the selection of classification features is highly demanding, and samples need to be re-selected for different data, resulting in low efficiency

[0008] In addition, the existing cloud detection methods still detect some clear sky, land, and ocean as clouds and some missed detections in satellite image cloud detection, and the cloud detection efficiency is low.

Images