Algorithm for correction of geometric distortions of spectral images recorded using an acousto-optical hyperspectrometer under conditions of angular deviations

Subject of study. The process of formation of spectral images by an acousto-optical hyperspectrometer under conditions of angular deviations of the equipment carrier is considered. Aim of study. This study aimed to develop a model of distortion of hyperspectral images generated by an acousto-optical hyperspectrometer under the conditions of angular oscillations of an unmanned aerial vehicle and synthesize an algorithm for correcting such distortions based on angular data from an inertial navigation system. Method. The research method is based on a photogrammetric model of the formation of spectral images under the conditions of angular deviations of the registration system. Main results. The relationship between the hyperspectrometer deviations at the angles of roll, pitch, and yaw and the nature of geometric distortions of the recorded spectral images is determined. An algorithm for correcting the distortions of spectral images based on the data of an inertial navigation system has been developed. The analysis of experimental and numerical studies of the developed algorithm showed a sufficiently high accuracy of correction of the geometric distortions of images caused by the angular fluctuations of the registration system in the range of possible measurement errors. Practical significance. The developed algorithm is computationally simple compared with the known ones and does not require an additional system for registering reference images or complex precalibration, which allows the correction of hyperspectral images on a time scale close to real time with acceptable quality. (c) 2024 Optica Publishing Group