Super-resolution imaging lidar based on Fourier ptychography

Abstract

Traditional imaging lidar exhibits an obvious trade-off between the resolution and the size of its optical system. In order to realize a miniaturized super-resolution (SR) imaging lidar, Fourier ptychography (FP) has been introduced to break through the diffraction limit of the camera lens. FP, derived from synthetic aperture method, is capable of acquiring high resolution and large field-of-view reconstructed images without increasing the aperture size by capturing multiple images with diverse incident angles before computationally combining with phase retrieval algorithm. In this work, a SR imaging lidar system was proposed by using reflective-type FP, which mainly consists of a s-CMOS camera, a Nd:YAG laser, and a 2-D translation stage so as to achieve aperture scanning on the x and y axes. To validate this technique experimentally, a set of images of a positive USAF chrome-on-glass target were obtained for quantitative analysis, and an uneven 1 yuan nickel-on-steel RMB coin was used to simulate the applicability of the SR imaging lidar in practical applications. The observations show that the obtained images based on FP technique have an obvious improvement in resolution, contrast, and clarity. It is worth mentioning that the resolution of these reconstructed images is increased over 3 times in the experiment on the USAF target. Moreover, the images under different apertures were collected, processed and analyzed, which suggest the initial image quality has a non-negligible influence on the reconstructed results. This technique not only improves the performance of the imaging lidar while maintaining low costs, but also bring new vitality in remote image recognition and analysis.