Section Two
Physical Limitations and How to Overcome Them
Even though CCD and CMOS imaging sensors are widely used and considered suitable for most imaging applications, they have various physical limitations, such as various sources of noise, limited dynamic range for sensing light intensity and finite spatial resolution. In addition, they have several intrinsic malfunctionings and the resulting artifacts such as smearing and blooming. Therefore, efforts to overcome these physical limitations have been made continuously since the initial invention of imaging sensors. In this section, the intrinsic physical limitations with the CCD and CMOS imaging sensors are described and solutions to overcome such problems are presented.
In section 2.1, various types of noise with imaging sensors which affect and degrade the quality of captured images are described. CCD and CMOS imaging sensors suffer from various sources of noise. Carnes and Kosonocky explained noise sources in CCD(#13). They classified the noises into four main categories, which are : transfer loss noise, background charge generation, output amplifier noise, and fast interface state trapping noise. Tompsett described and evaluated the limitation on transfer efficiency of interface states(#14). He suggested using a background charge in the device at all times for minimization of the limitation of transfer efficiency. Hopkinson and Lumb used correlated double sampling techniques to attenuate low frequency noises(#15). Burke and Gajar showed that an interface-state dark current can be dynamically suppressed by periodically exchanging charge packets between gates(#16). Centen examined the noise performance versus the dimensions of MOS transistor and suggested criteria for choosing the optimum gate dimensions(#17). Snyder et al. used two methods to compensate the readout noise of CCD(#18). One approximates the Gaussian noise by Poisson noise and applies the modified Richardson-Lucy algorithm, while the other is the expectation maximization algorithm. Frenkel et al. developed a model based on photon-noise-limited operation of intensified CCD cameras(#19) and Reich observed the Sub-Poisson statistics in an electronically shuttered and back-illuminated CCD(#20). Plichta and Milne reduced the dark current noise and restored black reference using correlated double sampling(#21). R
댓글 0
| 번호 | 제목 | 글쓴이 | 날짜 | 조회 수 |
|---|---|---|---|---|
| 6 | "CCD and CMOS Imagers" | srip | 2003.09.06 | 6403 |
| 5 | Table Of Contents | srip | 2003.09.06 | 9123 |
| 4 | Preface | srip | 2003.09.06 | 6012 |
| 3 | Introduction of Section One - Technical Overview | srip | 2003.09.06 | 7893 |
| » | Introduction of Section Two - Physical Limitations and How to Overcome Them | srip | 2003.09.06 | 8664 |
| 1 | Introduction of Section Three - Applications | srip | 2003.09.06 | 9155 |