Celestial image backgrounds are relevant to many upward-looking sensors, particularly those operating at altitudes where the presence of clouds is minimal, and the absorption of the atmosphere is negligible. In this paper, a method is presented for the generation of synthetic celestial background imagery. For the determination of star and planet signatures, the technique uses third party databases with the underlying assumption that the various spectra follow Planck’s blackbody spectral distribution. An Image Generation Tool (IGT) has been developed to create synthetic celestial imagery within any spectral band and for any imaging sensor. The approach taken within the IGT provides a fast and efficient approach for image generation which makes the technique viable for both single frame and image-sequence generation. Example image data is presented for the MWIR spectral band (3-5μm) and the effects of noise and image blur are examined. It is shown that the cooler, brighter stars can exceed practical noise-levels in the MWIR, leading to potential false-alarm signatures.
|