Optical Crosstalk in CMOS Image Sensors Chris Fesenmaier and Benjamin Sheahan |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Crosstalk Reduction - Conclusions Given the current trend of constantly decreasing pixel size, it is certain that spatial crosstalk will become a major issue for future generations of CMOS image sensors. In order to continue this size trend while maintaining image quality, optical design considerations must become more of a priority. The designs surveyed offer the promise of increased light sensitivity and decreased crosstalk, although each requires significant changes to a process that has been heavily optimized to reduce costs. As was expected, the best performers in this study, namely the high-index light guide and metal mirror designs, are the ones most difficult to fabricate. More research is necessary to identify new high-index materials and fabrication processes that would provide reliable improvements with little added cost. It is also important to consider the non-optical effects of these designs, which may include degraded mechanical and electrical properties. While it may appear from the lack of publications on this topic that crosstalk mitigation is not being aggressively pursued, the relative lack in public information about these techniques is misleading because companies are very protective of their future plans in such a competitive marketplace. It is expected that these crosstalk reduction techniques, or other novel ones, are currently being pursued and will be implemented in the near future. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||