Detecting the light intensity and transducing it to an electrical parameter (voltage or current), and subsequently processing the signals from an array of detectors, in spatial and/or temporal domain, are the primary tasks of vision chips. Understanding the physical and electronic principles playing role at each stage is of great importance in designing vision chips and developing systems based on vision chips. Although a significant amount of work has been done in this area, there is little literature concerning a systematic approach to the design of vision chips. The bulk of such a literature would consist of principles of phototransduction, spatial, and temporal processing. The present document tries to initiate and inspire such an effort and looks forward to a more established future for vision chips.
This chapter is organized as follows. In
section 7.2 the first elements of
phototransduction, i.e. photodetector devices are presented. Analytic
expressions for the quantum efficiency of different detector
structures in standard processes are derived in
section 7.2.1. Section 7.3
describes some of the circuits for phototransduction, or
photocircuits
.
Spatial processing principles, techniques, and circuits are discussed
in section 7.5. Temporal and spatio-temporal
processing methods are described in
section 7.6. Adaptation mechanism for
extending the dynamic range of the system are discussed in
section 7.7. Some of the problems, such as mismatch
and digital noise, in designing vision chips are addressed in
section 7.8. Finally, basic circuits and
design techniques for active pixel sensors (APS) are presented in
section 7.4.