Intelligent Selection Method of UAV Cognitive Navigation Adaptation Zone Based on Remote Sensing Imagery

With the rapid development of Unmanned Aerial Vehicle (UAV) technology, there is an increasing demand for efficient and reliable UAV visual cognitive navigation systems. Adaptation zone selection technology is a prerequisite for realizing UAV visual cognitive navigation. The traditional hierarchical rule-based adaptation zone selection method cannot extract the depth features of the image, has poor generalization, and suffers from a certain degree of mis-selection. Based on this, this paper proposes an intelligent selection method for UAV visual cognitive navigation adaptation zone based on remote sensing image depth features, which utilizes high-resolution satellite remote sensing image data to identify and select the most suitable area for UAV cognitive navigation and improves the autonomous flight capability and environmental adaptability of UAVs. The method first utilizes an inter-image transformation model to generate a simulated real-time map based on a satellite reference sub-map. Then the intelligent matching algorithm based on deep learning extracts the image depth features, and combined with the designed DADC mismatch rejection strategy, the matching results between the reference sub-map and the simulated real-time map are obtained. Finally, the selection of the adaptation zone is realized by calculating the average number of matches and the matching success rate. The experimental results show that the method proposed in this paper is effective and robust in different scenarios, which not only can effectively recognize the suitable area for UAV cognitive navigation intelligently from remote sensing data, but also provides a new method for UAV scene matching, and lays the foundation for UAV visual cognitive navigation under satellite denial conditions.