Generation of Simulated Satellite Images of CAS500-4 with Intentional Geometric Errors Using a Physical Sensor Model

This study proposes a simulation framework for generating satellite imagery using inverse orthorectification based on a physical sensor model and for evaluating whether intentionally injected geometric errors are reproduced in the simulated imagery.Since actual imagery is unavailable at the prelaunch stage, it is necessary to verify whether simulated imagery reflects the intended imaging geometry under prescribed satellite observation conditions.To address this, geometric errors were defined and injected at the physical parameter level using an orbit-attitude (OA) model.The proposed methodology was first validated using imagery from the currently operating KOMPSAT-3A satellite.The results showed that errors introduced into the position and attitude parameters were reproduced in the simulated imagery with the intended magnitudes and directional characteristics.The framework was then applied to CAS500-4.Although CAS500-4 has been launched, operational imagery was not available for this study.Simulated CAS500-4 imagery was generated using an OA physical model reflecting the satellite's design specifications and operational scenarios, followed by the same error insertion experiments.The results indicate that position errors, defined in terms of ground displacement, were consistently reproduced in image space, whereas attitude-induced displacements followed the expected geometric trends.These findings indicate that the proposed framework provides a controlled means of evaluating the propagation of prescribed geometric errors into simulated image geometry.The generated imagery can serve as test data for geometric-processing algorithms under defined error scenarios.