A Screening Method for Ground Object Features Oriented to High-Resolution Images

Abstract

The invention discloses a feature screening method for high-resolution images, which includes performing feature analysis and feature rough selection on target features according to the actual distribution of features and the expression form of remote sensing images; One-time feature extraction of the target features in the original image is carried out to obtain multiple extraction results of the target features; the constructed feature sensitivity model is used to perform sensitivity analysis on each extraction result, and the effectiveness of different image features on the target features is obtained. Extraction degree: sort the sensitivity of different image features, select a few features with high sensitivity to the target feature, and form a sensitive feature group of the target feature. Its remarkable effect is: considering multiple aspects comprehensively, it analyzes the sensitivity of features to target features, improves the objective authenticity of the extraction results, thus effectively solves the problem of difficult feature selection, and weakens the impact of human factors on ground features. Interference with information extraction.

Description

technical field [0001] The invention relates to the technical field of remote sensing image processing, in particular, to a method for screening feature features of high-resolution images. Background technique [0002] High-resolution remote sensing images contain rich detailed information of ground objects, complex and diverse spectral information, and rich structural information such as space, texture, topology, and semantics. It is inevitable that the feature data has a high dimensionality and a large number of phenomena, that is, The key to the effective extraction of ground object information is how to optimize the selection of massive ground object features due to the "dimension disaster". [0003] Existing feature selection methods are mainly researched from two aspects: search strategy and evaluation criteria. The former commonly used algorithms include exhaustive search, sequential search, genetic algorithm, ant colony algorithm, particle swarm algorithm, etc. The ...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the object of the present invention is to provide a feature screening method for high-resolution images, which constructs feature features from two elements: the accuracy of feature feature information extraction and the complexity of mis-extraction. Sensitivity analysis model, which quantifies the effective extraction degree of different image features for target features, and solves the problem of difficult feature selection

Images