MEMS-based super-resolution remote sensing system using compressive sensing

In this article we present a novel remote sensing system the resolution of which exceeds the image sensor's with the help of high-resolution MEMS device, which will reduce the cost of high-resolution image sensors. The super-resolution factors of 2 × 2 and 4×4 have been tested by simulation and remote sensing experiments. By applying the theory of compressive sensing, we can compress the acquisition data rate while preserving high image quality. We employ the IRLS as the super-resolution reconstruction method. A sample rate greater than 50% guarantees that our algorithm is more accurate. Through experiments we have found that the mismatch of the image sensor and MEMS device diminishes the image uniformity. Our algorithm embedded with non-uniformity correction greatly alleviates the problem visually and quantitatively.