Perceptual resolution improvements in far-field Fourier ptychography imaging

Abstract

Synthetic aperture techniques applied to visible imaging in the direction of rapid development in recent years and have achieved initial effectiveness in the fields of remote sensing. The existing synthetic aperture technique for visible imaging mainly relies on the application of Fourier Ptychography (FP) in the far field to obtain more high- frequency information by aperture scanning. FP allows resolution enhancement for the smooth object. However, the FP strategy is not sensitive to diffuse re action imaging formed by rough objects. The conventional approach to high-resolution reconstruction of images containing speckle noise by FP is to boost the overlap rate between sub-aperture images, which increases the amount of data required by a factor of ten and affects imaging efficiency. In this work, a synthetic aperture method via Total Variation (TV) regularization is proposed to achieve less- speckle imaging. The prior distribution is imposed on the object with the assistance of TV regularization and solved by the Augmented Lagrangian method. The reconstruction results of high quality without increasing data redundancy are obtained, and the optimal signal-to-noise ratio is achieved at a 70% overlap rate. The proposed method improves the resolution by a factor of eight and further enhances the perceptual resolution of rough objects. To the best of our knowledge, the findings achieve the best imaging quality without a high overlap rate, permitting large field-of-view, high-resolution detection of targets for widespread far-field detection and remote sensing applications.