We propose a new method for image noise detection and reduction in complementary metal oxide semiconductor (CMOS) image sensors inspired from audio noise cancelling techniques. Our algorithm is based on computing efficiently the time-dependent pixel auto correlation function (ACF) from constant time interval acquired sequences of images.
We demonstrate the effectiveness of our approach for successfully detecting and reducing white noise. Further, we consider an adaptive filter that exhibits significant computational improvements making it highly practical. Finally, we report on experiments displaying the high quality imaging systems obtained in practice.
Source: The Pennsylvania State University
Authors: Kazuhiro Hoshino | Toshihiro Nishimura