基于高光谱遥感的黄河滩区典型人工材质目标检测OA

The application of remote sensing technology for the timely and accurate monitoring of artificial target materials in the Yel-low River beach area is significant for ecological protection of the Yellow River.Hyperspectral object detection technology is an effec-tive tool for extracting typical materials.This study utilized a drone-captured hyperspectral imaging technology to systematically in-vestigate typical scattered targets in the Yellow River beach area.Twelve representative datasets,comprising three distinct back-ground categories and four target types,were constructed through comprehensive field surveys.Comparative experiments were con-ducted employing five object detection algorithms and five anomaly detection methodologies.The experimental results demonstrated that object detection algorithms exhibited significantly higher operational efficiency compared to most anomaly detection approaches.For known materials,a spectral database could be constructed to detect common garbage targets using object detection and direction-al material detection.Meanwhile,anomaly monitoring algorithms could be used to detect potential targets.