Smart CMOS Imaging Architectures

This book demonstrate two novel approaches of intelligent CMOS imaging using bio-mimicry. One approach is at the system level whereby pixel structure is the standard Active Pixel Sensor (APS) and the sampling architecture is changed from 1D row sampling to 2D ring sampling. The second approach, however, is based on a modified active pixel sensor structure while using standard row sampling. The first approach is coined Pyramidal CMOS image sensor whereas the second approach is coined a Multiresolution CMOS image sensor. At the same time, the two approaches present two Multiresolution approaches: one in the time domain and the other in the special domain. The outcome of the first design approach is a Foveated Dynamic Range (FDR) very similar to that of the human visual system whereas the outcome of the second design is a versatile multiresolution sampling. Both approaches share the same goal: Maximizing the sampled image information over the carrying sampled data.

Autorentext

Dr. Fayçal Saffih got his B.Sc in Solid­State Physics from University of Setif, Algeria, his M.Sc. in Artificial Intelligence (Physics) from University of Malaya, Malaysia, and his Ph.D. in Electrical & Computer Engineering from University of Waterloo, Canada. He is currently Assistant Professor at the American University of the MiddleEast, Kuwait.