Super-resolution reconstruction in a computational compound-eye imaging system

From consumer electronics to biomedical applications, device miniaturization has shown to be highly desirable. This often includes reducing the size of some optical systems. However, diffraction effects impose a constraint on image quality when we simply scale down the imaging parameters. Over the past few years, compound-eye imaging system has emerged as a promising architecture in the development of compact visual systems. Because multiple low-resolution (LR) sub-images are captured, post-processing algorithms for the reconstruction of a high-resolution (HR) final image from the LR images play a critical role in affecting the image quality. In this paper, we describe and investigate the performance of a compound-eye system recently reported in the literature. We discuss both the physical construction and the mathematical model of the imaging components, followed by an application of our super-resolution algorithm in reconstructing the image. We then explore several variations of the imaging system, such as the incorporation of a phase mask in extending the depth of field, which are not possible with a traditional camera. Simulations with a versatile virtual camera system that we have built verify the feasibility of these additions, and we also report the tolerance of the compound-eye system to variations in physical parameters, such as optical aberrations, that are inevitable in actual systems.