Method and system for identifying photovoltaic module in aerial image of unmanned aerial vehicle based on cross-domain detection

By applying cross-domain detection technology in drone aerial images and combining the information of satellite remote sensing images, the problem of difficult identification of distributed photovoltaic modules in drone images is solved, and higher recognition accuracy and recognition rate are achieved.

CN119693814APending Publication Date: 2025-03-25STATE GRID HUNAN ELECTRIC POWER COMPANY LIMITED +2
0 Cites 0 Cited by

Patent Information

Application Number
CN202411753902.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-02
Publication Date
2025-03-25

AI Technical Summary

Technical Problem

It is difficult for the prior art to accurately identify and locate distributed photovoltaic modules, especially in aerial images of drones, where there are problems of small data volume and low data quality.

Method used

Using a cross-domain detection method, drone aerial images are used as the target domain and satellite remote sensing images are used as the source domain. Cross-domain information is extracted through steps such as feature extraction, distance measurement, feature interaction and information fusion, and combined with prior guidance features, photovoltaic modules are identified through classification branches and regression branches.

Benefits of technology

It improves the recognition accuracy and recognition rate of photovoltaic modules in aerial images of drones, enhances the generalization ability and robustness of the model, and can more accurately locate photovoltaic modules.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a method and a system for identifying a photovoltaic module in an unmanned aerial vehicle aerial image based on cross-domain detection, and the method comprises the steps: taking the unmanned aerial vehicle aerial image and a satellite remote sensing image of the photovoltaic module as a target domain and a source domain, respectively selecting images from a target domain data set and a source domain data set, and extracting common features and non-common features; enabling common features to be close and non-common features to be far away through distance measurement, and enabling a classifier to act on the common features to strengthen the capability of distinguishing a source domain from a target domain by using the common features; fusing interaction feature information after the common features of the source domain and the target domain are interacted, calculating to obtain a priori enhanced target domain feature and a source domain feature according to a priori guidance feature and a fusion feature, and performing classification branch and regression branch on the priori enhanced target domain feature and the source domain feature to obtain a target domain feature; and carrying out threshold filtering and feature reduction to obtain photovoltaic module identification results of a source domain and a target domain. The method can accurately identify the photovoltaic module in the aerial image of the unmanned aerial vehicle, and has a wide application prospect.
Need to check novelty before this filing date? Find Prior Art