Abstract: Spaceborne optical remote sensing provides a novel approach for the accurate detection, positioning, and identification of flying aircraft. Due to the lack of techniques for extracting altitude information of flying aircraft from hyperspectral images, obtaining precise spectral reflectance of flying aircraft remains challenging, which hinders the analysis and recognition of their spectral characteristics. This thesis proposes a method for retrieving the altitude and spectral reflectance of flying aircraft from hyperspectral images based on contrails. By constructing a geometric model of contrails and their shadows, the altitude of flying aircraft is accurately retrieved. Furthermore, the spectral reflectance is retrieved through a radiative transfer model that couples the flying aircraft with the background environment. Experiments were conducted using GF-5 hyperspectral images in typical scenarios such as land, sea, and coastal backgrounds. The proposed method was validated by comparing the results with aircraft type and altitude information from Flightradar24, the FLAASH method for land surface spectral reflectance retrieval, and spectral reflectance data of parked/gliding aircraft in EnMAP hyperspectral images from typical international airports. The results demonstrate that the altitude retrieval accuracy for wide-body airliners in flight can reach ±100 m, and the spectral reflectance retrieval accuracy for flying aircraft is significantly improved compared to classical surface spectral reflectance retrieval methods. This study provides important support for the application of hyperspectral remote sensing images in the detection, positioning, and identification of flying aircraft.

Key words: hyperspectral remote sensing, flying aircraft, contrail, altitude, spectral reflectance