Analysis of self-correcting active pixel sensors

This paper evaluates the operation of self-correcting active pixel sensors presented in [6] using Signal-to-Noise Ratio. The evaluation is based on a simplified Active Pixel Sensing (APS) model. We show that in the absence of stuck faults (i.e., no errors) the performance of the system suffers from considerable degradation especially at low illumination (i.e., typical indoor scenes). We use the same model to quantify the number of defective pixels under which self correction is beneficial.

[1]  Eric R. Fossum,et al.  CMOS image sensors: electronic camera on a chip , 1995, Proceedings of International Electron Devices Meeting.

[2]  Glenn H. Chapman,et al.  Creating 35 mm camera active pixel sensors , 1999, Proceedings 1999 IEEE International Symposium on Defect and Fault Tolerance in VLSI Systems (EFT'99).

[3]  Israel Koren,et al.  A self-correcting active pixel camera , 2000, Proceedings IEEE International Symposium on Defect and Fault Tolerance in VLSI Systems.

[4]  Glenn H. Chapman,et al.  Design of a self-correcting active pixel sensor , 2001, Proceedings 2001 IEEE International Symposium on Defect and Fault Tolerance in VLSI Systems.

[5]  Abbas El Gamal,et al.  Pixel-level processing: why, what, and how? , 1999, Electronic Imaging.

[6]  Gerald C. Holst,et al.  CCD arrays, cameras, and displays , 1996 .

[7]  Eric R. Fossum,et al.  Active pixel sensors: are CCDs dinosaurs? , 1993, Electronic Imaging.