Alex Golts (EE, Technion)
Tuesday, 24.6.2014, 11:30
There is resolution loss that is caused strictly by noise, without the presence of imaging blur. Past studies analyze resolution limits in single channel, pan-chromatic systems. There is a need to generalize this analysis to color cameras. We analyze two main types of color sensors: Bayer sensors, which yield color by a filter array on top of a single detector array, and full-field systems (e.g. 3CCD). Bayer sensors compromise resolution. In order to quantify this inherent loss, we present a model for the spatial frequency response of a Bayer system. Our model deals with a color input signal, and accounts for both sampling and linear demosaicking. For full-field systems, we analyze the probability of resolving details as a function of spatial frequency, under noise. The analysis introduces theoretical bounds for performance. Color sampling indeed enhances the ability to recover objects under noise, given a fixed photon flux. Our framework assess the ability and probability of a color imaging system to distinguish an object of given size, color and SNR.
M.Sc research under the supervision of Prof. Yoav Schechner.