High-precision space camera self-calibration via structural priors and intelligent optimization.

During long-term on-orbit operation, space cameras are subjected to extreme thermal cycles, high vacuum, intense radiation, and complex mechanical loads, which cause drift in the geometric parameters of the imaging system. To improve the on-orbit camera's imaging quality without altering the optical system, existing calibration methods are insufficient. Therefore, this study proposes, to our knowledge, a novel on-orbit autonomous calibration method based on the structural features of the spacecraft. Unlike traditional ground-based calibration methods and satellite-based reference object-dependent approaches, this method innovatively utilizes the spacecraft's inherent solar panels as the calibration reference source and constructs a line feature detection model based on space-frequency domain collaborative analysis. By employing a dual-domain fusion strategy, sub-pixel-level extraction of the solar panel's structural line features is achieved. The optimal vanishing point is trained using abundant feature line segments, and a final error function is constructed. An intelligent optimization algorithm is then proposed to iteratively solve this optimization problem. Simulations and real-world experiments demonstrate that this method provides excellent stability and accuracy.