Improved phase curvature autofocus for stripmap synthetic aperture radar imaging

Abstract

Based on the theory of phase curvature autofocus (PCA) on stripmap synthetic aperture radar (SAR), an improved algorithm for increasing the accuracy of phase error compensation is presented in this study. PCA method was proposed to extend the phase gradient autofocus method for SAR systems in stripmap mode. The main problems concerned with the traditional PCA algorithm are related to selecting candidates in the image for phase error estimation, windowing, estimation procedure, and range shift due to the phase error. In this study, the modification of traditional PCA algorithm has been performed in different steps including the following: improving range‐compressed data, prominent points extraction, adaptive windowing, weighted maximum likelihood for phase error estimation, improving phase error result, range shift compensation, and determining the condition to end the iterations. Real data experiments demonstrate the success of the proposed autofocus method, which is applied to the stretched‐based pulsed mode SAR data set in the absence of highly accurate inertial navigation units.

Combined MAM-PCA autofocus for stripmap SAR

“…For stripmap SAR, nonparametric phase curvature autofocus (PCA) is commonly used [3]. As a slightly modified version of the phase gradient autofocus (PGA), PCA offers estimation of both low and high frequency motion errors, including quadratic and higher orders [4].…”

“…The MD-based approach is used to estimate unknown linear error in sub-aperture based PGA and extract residual RCM [8]. Modifications are made to PCA to improve its error estimation using Kalman filter [9] and extracting prominent points in the scene and applying adaptive window on them [3]. Weighted PCA is proposed to take the range-dependent nature of the phase error into account [2].…”