A ground object classification method based on combination of hyperspectral image and laser radar

CN121725355BActive Publication Date: 2026-06-23INNER MONGOLIA UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
INNER MONGOLIA UNIV OF TECH
Filing Date
2025-12-22
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Existing hyperspectral and lidar joint classification methods ignore modal sparsity differences, resulting in redundant and ambiguous feature representations. Multi-scale fusion lacks a competitive selection mechanism, making it difficult to adaptively focus on the most discriminative key scale in complex scenes.

Method used

By combining the multi-scale spectral-spatial sparse coding module (MS-SSSE) and the multi-scale geometric sparse coding module (MS-GSSE) with the competitive sparse selection module (CSS), global dependencies are captured and explicit scale competition relationships are established through the visual selective state space block (VSS Block) and the spatial competitive selection sub-module (SCS Block), thereby achieving modality-specific sparse modeling and multi-scale competitive selection.

Benefits of technology

It significantly improves the accuracy and robustness of land cover classification in complex scenarios, solves the problems of insufficient modality-specific expression and multi-scale feature redundancy, enhances the physical interpretability and discriminative power of features, and maintains excellent robustness under small sample conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121725355B_ABST
    Figure CN121725355B_ABST
Patent Text Reader

Abstract

The present application belongs to the technical field of remote sensing, and provides a ground object classification method based on combination of hyperspectral images and laser radar. The method comprises: acquiring hyperspectral images and laser radar data; labeling ground object categories in the processed data and making a label file; constructing a ground object detection model and training the model according to the data and the label file, the model comprising a multi-scale spectral-spatial sparse coding module for extracting hyperspectral sparse features, a multi-scale geometric structure sparse coding module for extracting laser radar geometric sparse features, and a competitive sparse selection module for multi-scale feature dynamic fusion; and finally generating a ground object classification result by using the trained model. The ground object classification method solves the problems of neglecting modal sparse differences and feature redundancy by means of a'modality-specific sparse modeling + multi-scale competitive selection' mechanism, and effectively improves the accuracy and robustness of ground object classification of multi-source remote sensing data in complex scenes.
Need to check novelty before this filing date? Find Prior Art