Architecture and Noise Analysis for Block-based Compressive Imaging

We study a new architecture for block-based compressive imaging. The architecture basically consists of a programmable aperture, multiple single-pixel sensors, and lens. The aperture is shared by all sensors which make the compressive measurement of a scene with different views. The combined measurements from the sensors can be used to reconstruct a unified image with a large number of pixels. This paper addresses two main issues present in compressive imaging: namely, the issue of long acquisition by using a single-pixel sensor, and that of high complexity due to the use of a very large sensing matrix for high resolution images. These issues are resolved in the new architecture of using block-based compressive sensing. We also show that the new architecture has better signal-to-noise ratio than other imaging architecture such as the single-pixel camera.