A Single-Pixel Imaging System for Remote Sensing by Two-Step Iterative Curvelet Thresholding

Abstract

Recently, a new framework named compressed sensing (CS) for the simultaneous sampling and compression of signals has been applied for panoramic-view imaging in aerospace remote sensing. By CS, it is possible for us to take superresolution photographs using only one or a few pixels rather than a million pixels by conventional digital cameras. However, the most popular approach of satellite/airborne remote sensing is line-scan imaging instead of panoramic-view imaging. In this letter, we propose a single-pixel imaging system for line-scan onboard cameras by applying compressive-scanning matrices in a sensing step and a two-step iterative curvelet thresholding method in an offline decoding step, which converges faster than previous single-step iterative thresholding methods. Numerical experiments show good performance of the proposed method for remote sensing. Results indicate the need to design practical single-pixel remote sensing instruments involving less storage space, less power consumption, and smaller size than the currently used charged-coupled-device cameras.