14-11-2012, 04:22 PM
The Comparison of CCD and CMOS Image Sensors
[attachment=39465]
ABSTRACT
The architectures of CCD and CMOS image sensors are introduced briefly, followed by comparison of their
performances in detail. At last, the future development trends of CCD and CMOS image sensors are prospected. It is
pointed out that CCD and CMOS image sensors will remain complementary and competition, and flourish the image
sensor market together in predictable future.
INTRODUCTION
CCD and CMOS image sensors are two main kinds of solid state image sensors which are widely applied at present.
CCD and CMOS image sensors used for photons detection are organized as arrays of photodetectors that deliver an
electrical signal related to the amount of photons that fall on the pixel surface during the integration time. They both use
the photoelectric effect in silicon, in either a photogate or a photodiode detector.
Since their invention in 1970[1], CCD image sensors have been the dominated the visible photon detection and image
capture[2] due to their high light sensitivity, low noise and high resolution, even though CMOS image sensors appeared in
1967. Because of the restriction of technologies, the research on CMOS image sensors gained less attention and
development. However, in the early 1990s, with the development of technologies of VLSI and signal processing, CMOS
image sensors have gained more attention [3~6] from many researchers and industries due to its low power dissipation,
low fabrication cost, compatibility with VLSI and radiation hardness. Now, several attempts are on the way to overcome
the disadvantages of CCD and CMOS image sensors in order to become the mainstream technology.
The architectures of CCD and CMOS image sensors are introduced briefly, followed by comparison of their
performances in detail. At last, the future development trends of CCD and CMOS image sensors are prospected. In
predictable future, CCD and CMOS image sensors will remain complementary and competition, and flourish the image
sensor market together.
The architecture of the CMOS image sensor
The architecture of the CMOS image sensor is shown in Fig.2. The main contribution of CMOS active pixel sensors is
the combination inside the pixel of the detector, the charge-to-voltage conversion and transistors providing buffering and
addressing capability.
PERFORMANCE COMPARISON
Sensitivity
Sensitivity is a measure of the ability of photosensitive unit to collect photons generated electrons. The light-sensitive
area of CCD in the pixel is larger than that of CMOS image sensor, which is decided by the architecture of the CCD. In
contrast, the pixel of CMOS image sensor is more complex than that of CCD, so the ratio of light-sensitive area to the
pixel’s total size is very small. Therefore, the sensitivity of CCD image sensor will be higher than that of CMOS image
sensor in the case of same pixel size.
Dynamic range
Dynamic range refers to the ratio of the pixel's saturation level to its signal threshold. In comparable circumstances, the
dynamic range of CCD is 2 times larger than that of CMOS image sensor. CCD enjoys significant noise advantages over
CMOS image sensor because of quieter sensor substrates (less on- chip circuitry), inherent tolerance to capacitance
variations and common output amplifiers with transistor geometries that can be easily adapted for minimal noise. CMOS
has bigger noise because of more the on- chip amplifier, the addressing circuit, the parasitic capacity and so on, in this
regard. The noise level of CMOS is not equivalent to CCD even if introduce external circuit for signal processing and
chip cooling. it can be seen that the CCD’s low noise characteristic is decided by its physical architecture.
Response uniformity
Uniformity is the consistency of response for different pixels under identical illumination conditions. Ideally, behavior
would be uniform, but spatial wafer processing variations, particulate defects and amplifier variations create
nonuniformities. It is important to make a distinction between uniformity under illumination and uniformity at or near
dark. CMOS image sensor was traditionally much worse under both regimes. Each pixel have an open- loop output
amplifier and the offset and gain of each amplifier varied considerably because of wafer processing variations, making
both dark and illuminated nonuniformities worse than those in CCD. However, feedback-based amplifier architectures
can trade off gain for greater uniformity under illumination. The amplifiers have made the illuminated uniformity of
some CMOS image sensors closer to that of CCD. CMOS manufacturers make great efforts to low dark nonuniformity,
but dark nonuniformity caused by the offset of amplifier is worse than that of CCD. This is a significant issue in high –
speed applications, where limited signal levels mean that dark nonuniformities contribute significantly to overall image
degradation.
Quantum efficiency
CCD and CMOS detectors share several common aspects in terms of quantum efficiency (QE) and both are being
manufactured in monocrystalline silicon, using photoelectric effect for electron–hole pair generation and electric field for
carrier separation. Due to different fabrication targets, CCD and CMOS devices differ strongly in both composition and
thickness of top layers, and space charge region depth and doping. Incident visible photons in both devices have to go
through a stack of top layers before being absorbed by silicon. Due to the different refractive indexes (Si:3–5;
SiO2:1.45), thickness and nature of the various materials used in this stack, transmission will be limited and wavelength
dependent. The CCD top-level architecture is in any case simpler and can be optimized for high transmission value.
However, remarkable improvements in the transmission losses of CMOS detector have been obtained, for example
through the use of near-surface antireflection layers, by CMOS foundries interested in the imaging market that have
developed dedicated process modules for high performance detector compatible with the core mixed signal process[7].
DEVELOPMENT TRENDS
CCD is broadly applied and its technology is now very mature. At present, it has turned toward large-size arrays in terms
of both pixel number and area, small pixel size, ultraviolet spectral response, in order to adapt the need of digital camera,
digital vidicon, scanner and other science fields. CCD technology will continue to dominate the high- performance
branch [14]. Spirit and Opportunity have successfully landed on Mars in 2004, on which high definition panoramic camera
adopt 4000 × 4000 pixels CCD image sensor which developed by Canada DALSA corporation. Lockheed Matin
Corporation developed a 9000×9000 pixels CCD image sensor with 8.75μm×8.75μm pixel size in 1996. A 10000×10000
pixels CCD with 10μm×10μm pixel size had been designed in 2000. Sanyo Corporation developed a CCD image sensor
which pixel size decreases to 2.5μm×2.5μm and corresponding space resolution reaches 200lp/ mm.
At present, CMOS image sensor has turned toward high resolution, high dynamic range, high sensitivity, micromation,
digital and multifunction [15]. Foveon Inc. developed a 4096×4096 pixels CMOS image sensor in 2000. Micron
technology developed MT9E001, a 1/2.5 inch 8 mega pixels CMOS image sensor with 1.75μm×1.75μm pixel size in
March 2007. CMOS image sensor will be widely used along with the perfection and development of its technology. The
global sales volume of CMOS image sensor will greatly increase year by year. According to the latest research report of
IC Insights, the global sales income of CMOS image sensors in 2008 will increase by 19% compared With 2007, and
reach 4.4 billion dollars. IC Insights estimates the sales volume of CMOS image sensors accounts for 58% of the whole
image sensors in 2008, while that is 53% and 46% in 2007 and 2004 respectively. It is predicted that the sales volume of
CMOS image sensors will accounts for 73% of the whole image sensors in 2012.
CONCLUSIONS
Through the development for the last three decades, CCD technology, due to process specialization, has been able to
provide top-level performances for detection but at the cost of both several drawbacks for the user and, for the best
products thinned and backside illuminated, the risk associated with the very limited number of procurement sources. It
remains the first choice technology for very high-end applications, such as medical imaging, astronomy, low-end
professional cameras, etc. because of its better image quality. On the other hand, CMOS image sensors have intrinsic
advantages (low power consumption, low cost, high speed imaging, integration capability, radiation hardness etc.) that
make them well suited not only for low-cost imaging markets but also for high performances applications such as highend
Digital Still Photography, High-Definition Television and several space applications. In predictable future, CCD and
CMOS image sensors will remain complementary and long term competition, and flourish the image sensor market
together.
[attachment=39465]
ABSTRACT
The architectures of CCD and CMOS image sensors are introduced briefly, followed by comparison of their
performances in detail. At last, the future development trends of CCD and CMOS image sensors are prospected. It is
pointed out that CCD and CMOS image sensors will remain complementary and competition, and flourish the image
sensor market together in predictable future.
INTRODUCTION
CCD and CMOS image sensors are two main kinds of solid state image sensors which are widely applied at present.
CCD and CMOS image sensors used for photons detection are organized as arrays of photodetectors that deliver an
electrical signal related to the amount of photons that fall on the pixel surface during the integration time. They both use
the photoelectric effect in silicon, in either a photogate or a photodiode detector.
Since their invention in 1970[1], CCD image sensors have been the dominated the visible photon detection and image
capture[2] due to their high light sensitivity, low noise and high resolution, even though CMOS image sensors appeared in
1967. Because of the restriction of technologies, the research on CMOS image sensors gained less attention and
development. However, in the early 1990s, with the development of technologies of VLSI and signal processing, CMOS
image sensors have gained more attention [3~6] from many researchers and industries due to its low power dissipation,
low fabrication cost, compatibility with VLSI and radiation hardness. Now, several attempts are on the way to overcome
the disadvantages of CCD and CMOS image sensors in order to become the mainstream technology.
The architectures of CCD and CMOS image sensors are introduced briefly, followed by comparison of their
performances in detail. At last, the future development trends of CCD and CMOS image sensors are prospected. In
predictable future, CCD and CMOS image sensors will remain complementary and competition, and flourish the image
sensor market together.
The architecture of the CMOS image sensor
The architecture of the CMOS image sensor is shown in Fig.2. The main contribution of CMOS active pixel sensors is
the combination inside the pixel of the detector, the charge-to-voltage conversion and transistors providing buffering and
addressing capability.
PERFORMANCE COMPARISON
Sensitivity
Sensitivity is a measure of the ability of photosensitive unit to collect photons generated electrons. The light-sensitive
area of CCD in the pixel is larger than that of CMOS image sensor, which is decided by the architecture of the CCD. In
contrast, the pixel of CMOS image sensor is more complex than that of CCD, so the ratio of light-sensitive area to the
pixel’s total size is very small. Therefore, the sensitivity of CCD image sensor will be higher than that of CMOS image
sensor in the case of same pixel size.
Dynamic range
Dynamic range refers to the ratio of the pixel's saturation level to its signal threshold. In comparable circumstances, the
dynamic range of CCD is 2 times larger than that of CMOS image sensor. CCD enjoys significant noise advantages over
CMOS image sensor because of quieter sensor substrates (less on- chip circuitry), inherent tolerance to capacitance
variations and common output amplifiers with transistor geometries that can be easily adapted for minimal noise. CMOS
has bigger noise because of more the on- chip amplifier, the addressing circuit, the parasitic capacity and so on, in this
regard. The noise level of CMOS is not equivalent to CCD even if introduce external circuit for signal processing and
chip cooling. it can be seen that the CCD’s low noise characteristic is decided by its physical architecture.
Response uniformity
Uniformity is the consistency of response for different pixels under identical illumination conditions. Ideally, behavior
would be uniform, but spatial wafer processing variations, particulate defects and amplifier variations create
nonuniformities. It is important to make a distinction between uniformity under illumination and uniformity at or near
dark. CMOS image sensor was traditionally much worse under both regimes. Each pixel have an open- loop output
amplifier and the offset and gain of each amplifier varied considerably because of wafer processing variations, making
both dark and illuminated nonuniformities worse than those in CCD. However, feedback-based amplifier architectures
can trade off gain for greater uniformity under illumination. The amplifiers have made the illuminated uniformity of
some CMOS image sensors closer to that of CCD. CMOS manufacturers make great efforts to low dark nonuniformity,
but dark nonuniformity caused by the offset of amplifier is worse than that of CCD. This is a significant issue in high –
speed applications, where limited signal levels mean that dark nonuniformities contribute significantly to overall image
degradation.
Quantum efficiency
CCD and CMOS detectors share several common aspects in terms of quantum efficiency (QE) and both are being
manufactured in monocrystalline silicon, using photoelectric effect for electron–hole pair generation and electric field for
carrier separation. Due to different fabrication targets, CCD and CMOS devices differ strongly in both composition and
thickness of top layers, and space charge region depth and doping. Incident visible photons in both devices have to go
through a stack of top layers before being absorbed by silicon. Due to the different refractive indexes (Si:3–5;
SiO2:1.45), thickness and nature of the various materials used in this stack, transmission will be limited and wavelength
dependent. The CCD top-level architecture is in any case simpler and can be optimized for high transmission value.
However, remarkable improvements in the transmission losses of CMOS detector have been obtained, for example
through the use of near-surface antireflection layers, by CMOS foundries interested in the imaging market that have
developed dedicated process modules for high performance detector compatible with the core mixed signal process[7].
DEVELOPMENT TRENDS
CCD is broadly applied and its technology is now very mature. At present, it has turned toward large-size arrays in terms
of both pixel number and area, small pixel size, ultraviolet spectral response, in order to adapt the need of digital camera,
digital vidicon, scanner and other science fields. CCD technology will continue to dominate the high- performance
branch [14]. Spirit and Opportunity have successfully landed on Mars in 2004, on which high definition panoramic camera
adopt 4000 × 4000 pixels CCD image sensor which developed by Canada DALSA corporation. Lockheed Matin
Corporation developed a 9000×9000 pixels CCD image sensor with 8.75μm×8.75μm pixel size in 1996. A 10000×10000
pixels CCD with 10μm×10μm pixel size had been designed in 2000. Sanyo Corporation developed a CCD image sensor
which pixel size decreases to 2.5μm×2.5μm and corresponding space resolution reaches 200lp/ mm.
At present, CMOS image sensor has turned toward high resolution, high dynamic range, high sensitivity, micromation,
digital and multifunction [15]. Foveon Inc. developed a 4096×4096 pixels CMOS image sensor in 2000. Micron
technology developed MT9E001, a 1/2.5 inch 8 mega pixels CMOS image sensor with 1.75μm×1.75μm pixel size in
March 2007. CMOS image sensor will be widely used along with the perfection and development of its technology. The
global sales volume of CMOS image sensor will greatly increase year by year. According to the latest research report of
IC Insights, the global sales income of CMOS image sensors in 2008 will increase by 19% compared With 2007, and
reach 4.4 billion dollars. IC Insights estimates the sales volume of CMOS image sensors accounts for 58% of the whole
image sensors in 2008, while that is 53% and 46% in 2007 and 2004 respectively. It is predicted that the sales volume of
CMOS image sensors will accounts for 73% of the whole image sensors in 2012.
CONCLUSIONS
Through the development for the last three decades, CCD technology, due to process specialization, has been able to
provide top-level performances for detection but at the cost of both several drawbacks for the user and, for the best
products thinned and backside illuminated, the risk associated with the very limited number of procurement sources. It
remains the first choice technology for very high-end applications, such as medical imaging, astronomy, low-end
professional cameras, etc. because of its better image quality. On the other hand, CMOS image sensors have intrinsic
advantages (low power consumption, low cost, high speed imaging, integration capability, radiation hardness etc.) that
make them well suited not only for low-cost imaging markets but also for high performances applications such as highend
Digital Still Photography, High-Definition Television and several space applications. In predictable future, CCD and
CMOS image sensors will remain complementary and long term competition, and flourish the image sensor market
together.