Axial Attitude Estimation of Spacecraft in Orbit Based on ISAR Image Sequence

Attitude estimation is a significant task for monitoring noncooperative spacecraft in orbit. Inverse synthetic aperture radar (ISAR) is an effective method to obtain the structure, size, and attitude of spacecraft. Most existing methods of attitude estimation based on ISAR images generally rely on characteristic point matching or rectangular feature matching among ISAR images. They often need high quality ISAR images, so the robustness is usually poor. A novel method for estimating spacecraft's axial attitude in ISAR image is proposed in this article. This method does not need to utilize a simulation database, or to match sequence of specific features, so it has good robustness. Our method bridges Range-Doppler images and attitude parameters with the accommodation of spacecraft's orbit parameters and the ISAR geometric projection model. Based on the ISAR imaging turntable model, the accurate position of the imaging plane of the spacecraft in orbit is derived. Then, attitude parameters are estimated by an optimization of the projection of spacecraft's three-dimensional model to the imaging plane. Experimental results demonstrate the effectiveness and robustness of the proposed method to estimate attitude of spacecraft in orbit by the simulated and measured ISAR images.