A signal-processing based method for acquiring very high resolution images with multiple cameras and its theoretical analysis

Proposes a signal-processing based method for acquiring a higher resolution image by processing multiple low resolution images taken with different cameras. The method consists of registration and reconstruction. The theoretical analysis based on the aperture effect and the projection filter principle shows that the upper bound in the performance of the proposed image acquisition method depends on the aperture of an imager. Experimental simulations are performed and the resolution improvements are examined. The further discussions are; investigation of registration methods that are robust even under severe aliasing effects, the configuration and arrangement of the multiple imagers, trade-off relationship between resolution improvements and SNR improvements when multiple images are used, the application to real three dimensional scenes etc