Super-resolution reconstruction algorithm using vibro-imaging and correlation image sensor

摘要

In this paper, we propose a novel super-resolving imaging scheme using vibratory optics and correlation image sensor. Human eyes always vibrate slightly during fixation (visual gaze on an identical point), which is called involuntary eye movements and is considered to play an important role in visual perception. Mimicking this phenomenon, our vibro-imaging system makes an image vibrate in a circular locus using a biaxial vibration mirror in front of the lens. As a result, the information about spatial intensity distribution around each pixel is modulated into the time-varying incident light on the pixel. A correlation image sensor (CIS) demodulates it using reference signals synchronized with the image vibration. This system enables us to obtain three images (intensity image and complex correlation image) with three different spatial frequency characteristics. From them, we show the frequency components over Nyquist frequency can be restored using only a single frame. Several numerical simulations and experimental results display the effectiveness of the scheme.