Remote Sensing Image Classification Method Based on Partially Random Supervised Discrete Hashing

Abstract

The invention discloses a remote sensing image target classification method based on partial random supervised discrete hash. According to the calibrated real data, target segmentation is performed on the remote sensing image, and feature extraction is performed on the segmented target. Each target is represented by a feature vector, and the feature vectors of all targets are combined into a feature matrix; The ratio is divided into training samples and test samples; discrete hash coding is performed on all samples; partial random hash coding is performed on all samples; discrete hash coding is combined with partial random coding, and parameters are iteratively optimized to finally obtain a more accurate hash Hash coding; according to the generated hash code, calculate the Hamming distance and complete the classification. This method solves the problem of high computational complexity caused by excessive data volume in the process of processing remote sensing images, and realizes fast and effective classification of remote sensing images.

Description

technical field [0001] The invention relates to a remote sensing image classification method, in particular to a remote sensing image classification method based on partial random supervised discrete hash. Background technique [0002] Due to the rapid development of satellite and aircraft technology, the application of remote sensing data has become more and more extensive, and target classification has gradually become one of the most important tasks in remote sensing data analysis. However, with the significant increase in data volume and resolution of remote sensing images, object classification has also become more challenging. Therefore, an effective feature representation method is very meaningful for remote sensing image target classification. In recent years, many technologies in this area have been proposed, which can be roughly divided into three categories: methods based on manual features, methods based on deep feature learning, and methods based on unsupervise...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a remote sensing image classification method, aiming at the problem of large amount of remote sensing image data and complex calculation, combined with data independent and data dependent methods, to complete the accurate classification of remote sensing image hash representation

Images