A Target Echo Envelope Alignment Method Integrating Historical Image Correlation and Radar Guidance Information

Envelope alignment is the first step in the translational compensation process of Inverse Synthetic Aperture Radar (ISAR) imaging. The accuracy of envelope alignment has a significant impact on azimuth focusing and imaging quality. However, translational motion prediction errors can lead to range offsets and phase errors, necessitating envelope alignment before azimuthal imaging. Traditional envelope alignment methods face challenges when extended observation times reduce pulse correlation while deep learning-based methods, though offering improved alignment performance, often rely on consistent electromagnetic scattering models and are typically non-causal, limiting their robustness and real-time applicability. To address these challenges, this paper proposes an envelope alignment method that leverages guiding information to restore one-dimensional range profiles and calculates offsets using the entropy and correlation of historical echoes. The proposed method achieves higher alignment performance to optimal traditional and machine learning-based methods while preserving echo details, especially in extended observation scenario. It is computationally efficient, operating synchronously with the generation of one-dimensional range profiles, and does not rely on future information. By integrating guiding information and historical correlation fusion, the method enhances alignment robustness and is suitable for real-time radar tracking systems.