A remote sensing data processing method based on multi-granularity

Abstract

The invention relates to a remote sensing data processing method based on multi-granularity, comprising the following steps: 1, uniformly integrating remote sensing through time order, and adding labels according to the data sensor source; 2, inputting that expert rule data as a granulation standard, and performing granulation compute screening and classification; 3, counting the label, dividing coarse grain data accord to that number of labels; 4, screening the data in the step 3 by using the existing clear remote sensing data as a rule; 5, screening that secondary granule calculation screening classification data obtain from the step 4, transferring the screening rules in the step 4, and carrying out repeated screening without changing the data; 6, modifying that screening rule in the step 5 to facilitate the next data screening according to the characteristics of the fine-grained data. A large number of existing remote sensing data are divided into classification information based on grain classification method, and all the remaining valuable information is obtained by grain calculation after the useful remote sensing information is found, which is applied to other classification of the same kind of data to facilitate the processing of subsequent data.

Description

technical field [0001] The invention relates to a remote sensing data processing method, in particular to a multi-granularity-based remote sensing data processing method. Background technique [0002] In the past ten years, high spatial resolution remote sensing images have been widely used in agriculture, forestry, ocean and environmental monitoring and other fields, and have huge economic value and social benefits. However, due to the large volume of high-spatial-resolution remote sensing images, various data types, rich information, and complex interpretation and analysis processes, it has been difficult to accurately and efficiently automatically perform high-spatial-resolution remote sensing images. classification of features. How to classify ground objects on high spatial resolution remote sensing big data has become one of the technical difficulties and bottlenecks affecting its large-scale application. Compared with medium and low resolution remote sensing images, ...

AI Technical Summary

Problems solved by technology

[0003] The present invention provides a multi-granularity-based remote sensing data processing method for the large amount of existing remote sensing data and it is difficult to find useful information. It can split a large number of existing remote sensing data into classified information based on granular classification methods , and after discovering useful remote sensing information, obtain all the remaining valuable information through granular computing, and apply it to other classifications of similar data according to algorithm statistics, so as to facilitate subsequent data processing