Method and system for establishing and verifying remote sensing image classification model, and electronic equipment

Abstract

The invention discloses a remote sensing image classification model establishment and verification method and system and electronic equipment. The method comprises the following steps: 1, acquiring classification sample data of a remote sensing image; 2, reading coordinates and categories of classified sample image data; 3, calculating the shortest distance between any two pieces of sample image data under each category; 4, comparing a calculation result with a threshold value, and judging whether the sample image data are processed in the same way or not according to a comparison result; 5, obtaining total sample image data after all the sample image data are identical, and randomly distributing the total sample image data according to a preset proportion to obtain a training sample set and a verification sample set; 6, establishing a model according to the training sample set, and verifying the model according to the verification sample set. According to the method, the spatial autocorrelation between the training sample and the verification sample can be removed, the objectivity and accuracy of remote sensing classification result verification are ensured, and the precision of the remote sensing classification result is not excessively evaluated.

Description

technical field [0001] The invention relates to the field of remote sensing images, in particular to a remote sensing image classification model establishment and verification method, system and electronic equipment. Background technique [0002] Supervised classification is one of the main methods of remote sensing classification. A necessary condition for supervised classification is to prepare remote sensing sample data, and the quality of remote sensing sample data is the key to achieve high-precision remote sensing classification. Usually, after a set of remote sensing sample data is prepared, a random allocation method is used to divide all samples into training samples and verification samples according to a certain proportion, such as 70% / 30%, and then an appropriate classification algorithm is used to construct training samples. The classification model is used to classify the entire remote sensing image, and finally the verification sample is used to test and evalu...

AI Technical Summary

Problems solved by technology

Using a simple random separation method, adjacent pixels are often assigned to training samples and verification samples respectively. The verification results obtained by using such verification samples are often falsely high, and accurate and objective results cannot be given.

Images