Measurement geometry strategies for super-resolution image reconstruction with multiple steerable sub-imagers

Using a set of low resolution images with sub-pixel shifts to reconstruct a high resolution less aliased image requires both interleaving of the image samples at the effectively higher sampling rate and deconvolution of the blur introduced by pixel sensor averaging. When measurement noise is low and knowledge of sub-pixel shift values is accurate, resolution improvement is limited primarily by the number of low resolution shifted images. However, at noise levels characteristic of typical image sensors, additional low resolution images are needed for noise reduction. If the relative shifts from a steerable array are not known precisely, estimating the shifts of the aliased images is required, and errors in shift positions may cause significant reconstruction errors. This paper considers the effects of both noise and position estimation errors when the temporal acquisition rate for the low resolution images is higher than the needed production rate of high resolution images so an over-determined image set can be used to reduce noise. In this context implications of measurement strategies on interpolation and restoration components of reconstruction are explored.