A charge-coupled device (CCD) is an integrated circuit with an array of capacitors that can detect light and convert the image into digital signals. By controlling external circuits, each small capacitor transfers its charge to an adjacent capacitor. CCDs are widely used in digital photography, astronomy, particularly in photometry, optical and spectroscopic telescopes, and high-speed imaging technologies like lucky imaging.
CCD was invented in 1969 by Willard Boyle and George E. Smith at Bell Labs in the United States. At the time, Bell Labs was working on video telephony and semiconductor bubble memory. By combining these two new technologies, Boyle and Smith developed a device they named the "Charge 'Bubble' Device." The characteristic of this device was that it could transfer charge along the surface of a semiconductor, initially attempting to use it as a memory device, where the charge could only be injected into the memory from a register. However, they soon discovered that the photoelectric effect could generate charges on the device surface, creating a digital image.
By 1971, researchers at Bell Labs could already capture images using a simple linear device, thus leading to the birth of CCD. Several companies followed up on this invention and started further research, including Fairchild Semiconductor, RCA, and Texas Instruments. Fairchild Semiconductor was the first to bring products to market, releasing a 500-element linear device and a 100x100 pixel array device in 1974.
In January 2006, Boyle and Smith were awarded the Charles Stark Draper Medal by the IEEE to recognize their contribution to the development of CCD. In October 2009, the two also received the Nobel Prize in Physics.
CCD sensors with grid-arranged pixels are used in digital cameras, optical scanners, and camcorder imaging sensors. Their light efficiency reaches 70% (capturing 70% of the incident light), superior to traditional film's 2%, which quickly led to the widespread adoption of CCDs among astronomers.
After an image is focused onto the surface of the capacitor array by a lens, it forms charges of varying strength on each capacitor unit depending on brightness. The linear CCD used in fax machines or scanners captures a narrow strip of light at a time, whereas the planar CCD used in digital cameras or camcorders captures an entire image or extracts a square area. Once exposure is complete, control circuits transfer the charge on the capacitor units to the next adjacent unit. When reaching the edge of the array, the electrical signal enters an amplifier, converting it to voltage. This process continues until the entire image is converted into voltage, sampled, digitized, and stored in memory. Stored images can be sent to printers, storage devices, or displays. Cooled CCDs were also widely used in astrophotography and various night vision devices in the early 1990s, with major observatories continually developing high-pixel-count CCDs to capture extremely high-resolution celestial images.
CCDs have a unique application in astronomy that allows fixed telescopes to act as if they have tracking capabilities. This is achieved by synchronizing the charge transfer on the CCD with the motion of the celestial object, thus enabling CCD guiding. Not only does this effectively correct tracking errors, but it also allows the telescope to record a larger field of view than originally possible.
Most CCDs are sensitive to infrared light, which has led to infrared imaging, night vision devices, and zero-illumination (or near-zero-illumination) cameras. To reduce infrared interference, astronomical CCDs are often cooled with liquid nitrogen or semiconductor coolers since objects at room temperature emit blackbody radiation in the infrared spectrum. CCD sensitivity to infrared also results in another effect—various CCD-equipped digital cameras or camcorders without infrared filters easily capture infrared light emitted by remote controls. Lowering the temperature reduces dark current in the capacitor array, increasing CCD sensitivity at low illumination levels, and improving sensitivity to ultraviolet and visible light (signal-to-noise ratio increases).
Temperature noise, dark current, and cosmic radiation can affect the pixels on the CCD surface. Astronomers use the shutter to open and close the CCD for multiple exposures, averaging the results to mitigate interference effects. To remove background noise, an average value of the image signal is first taken with the shutter closed, known as a "dark frame." After opening the shutter, the image is obtained, and the dark frame value is subtracted before filtering out system noise (dark spots, bright spots, etc.) to achieve clearer details.
The cooled CCD cameras used in astrophotography must be mounted in position to prevent external light or vibration interference. Since most imaging platforms are inherently bulky, astronomers use "auto-guiding" to capture faint celestial objects like galaxies or nebulae. Most auto-guiding systems use an additional off-axis CCD to monitor any shifts in the image, but some systems connect the primary mirror directly to the CCD camera used for photography. With an optical device that directs part of the starlight from the main mirror into a second CCD guide sensor, it can quickly detect minor tracking errors and automatically adjust the drive motor to correct them without needing additional guiding devices.
Most color digital cameras use a Bayer filter mounted on the CCD. Each group of four pixels forms a unit: one pixel filters red, one blue, and two green (since human eyes are more sensitive to green). As a result, each pixel receives a light signal, but the color resolution is lower than the light sensitivity resolution.
A 3CCD system using three CCDs and a beam-splitting prism can separate colors better. The prism splits the incident light into three colors—red, blue, and green—which are then captured by three separate CCDs. All professional-grade digital camcorders and some semi-professional camcorders use 3CCD technology.
As of 2005, ultra-high-resolution CCD chips remain quite expensive, and high-resolution still cameras equipped with 3CCD often exceed the budget of many professional photographers. Therefore, some high-end cameras use rotating color filters to achieve both high resolution and faithful color reproduction. These types of multi-shot cameras are only suitable for photographing static subjects.
In recent years, it has become possible to produce practical active pixel sensors (APS) using complementary metal-oxide-semiconductor (CMOS) technology. CMOS is the mainstream technology for all silicon chip manufacturing, and CMOS sensors are not only inexpensive to produce but also integrate signal processing circuits into a single device. This feature helps filter out background noise because CMOS is more prone to noise interference than CCD. However, this issue has gradually been resolved, thanks to the use of low-level amplifiers for individual pixels instead of a single high-level amplifier for the entire CCD array. Compared to CCDs, CMOS sensors consume less power and transmit data faster. CMOS sensors are more commonly seen in high-resolution digital video cameras and digital still cameras, especially in larger format digital single-lens reflex cameras, and consumer digital cameras have also begun using back-illuminated CMOS to improve image quality.
The invention of the CCD led to Willard Boyle and George Smith, along with Charles Kao, the inventor of fiber optics, sharing the 2009 Nobel Prize. The Nobel Prize committee stated that their inventions helped lay the foundation for today's networked world, creating many innovations for everyday life and providing tools for scientific advancement.
蘊藏許多助人的知識與智慧。