Deep Learning-Driven Attitude Jitter Estimation
摘要
Imagine a photographer trying to capture a crystal-clear portrait while standing on a gently swaying boat—the resulting image would inevitably suffer from blur and distortion. This analogy captures the essence of one of the most persistent challenges in space-based Earth observation: platform jitter. Like an unsteady hand holding a camera, even the slightest vibrations in a satellite’s structure can transform what should be razor-sharp imagery into blurred, geometrically distorted data that compromises scientific measurements and operational applications.