Lucent's Advanced "Camera On A Chip" Silicon Technology Promises Widespread Applications

MURRAY HILL, N.J. -- Vanguard International Semiconductor Corporation has licensed high-performance "camera on a chip" technology from Lucent Technologies' Bell Labs. The technology will lead to marble-sized video cameras -- ideal for PC videoconferencing and security cameras -- because each camera will use a single quarter-inch silicon chip.

The "camera on a chip" approach, based on the same CMOS (complementary metal oxide semiconductor) technology found in today's computer chips, produces real-time video images that rival the quality of images produced by camcorders, which rely on a handful of non-CMOS chips.

"While other companies have tried to use a single CMOS chip in video cameras, we're the first group to show high-quality performance," said researcher Bryan Ackland of Bell Labs, which is the research and development arm of Lucent Technologies.

The key to the CMOS camera's high performance is how Bell Labs researchers improved upon state-of-the-art "active pixel" technology, and this expertise complements Vanguard's ability to produce low-cost silicon chips. "We see an opportunity in CMOS technology, and we're fortunate to have Lucent as our partner," said Bob Swartz, director of Vanguard's Imager Product business unit.

Third-party camera and computer peripheral manufacturers will package the Vanguard chips with a small lens, and the final camera eventually will sell for less than $50, Swartz said.

By developing a high-quality imaging array using conventional CMOS technology, the researchers integrated all of the functions normally associated with a camera -- timing and control, analog to digital conversion and the signal processing required to provide exposure control and color balance -- onto a single silicon chip.

Traditonally, video cameras have required more chips because the camera relies on a specialized imaging process, known as CCD (charge-coupled device) technology, which was developed at Bell Labs in 1969. It turns out that the CCD manufacturing process is not well suited for making the different types of transistors necessary for the various non-image capturing functions.

The imaging array on the Bell Labs chip is comprised of more than 100,000 optical sensors (or pixels) laid out in a two-dimensional grid on the silicon surface. Each pixel generates a small packet of charge when illuminated by an image. Individual pixels are accessed using a two-dimensional arrangement of address and data buses, which is similar to how semiconductor memories are accessed. A small amplifier at each pixel helps reduce noise and distortion levels.

"Although the 'active pixel' approach was first proposed back in the early 70s," said Marc Loinaz, a researcher at Bell Labs, "it's only been in the last few years that technology has progressed enough where transistors can be included in each pixel without significantly increasing the size of the array."

Even with this change, however, CMOS-based cameras have been plagued by a defect known as "fixed pattern noise." This is an annoying stationary background pattern in the image that results from small differences in the behavior of the individual pixel amplifiers. Although some researchers believed this defect would prevent CMOS from ever seriously challenging CCDs, the researchers developed circuits outside the sensor array that detect and cancel this noise.

Besides requiring less space, CMOS cameras use less power than CCD cameras. A nine-volt battery, for instance, powers a CMOS camera for five hours, but powers one of today's computer-based desktop cameras for only 30 minutes. As a result, CMOS cameras are well suited for hand-held cameras or security cameras.

The CMOS camera also allows users to immediately access specific portions of an image. With today's video camera, users first must produce the entire image before manipulating it.

Although the new technology initially will be used for computer and security cameras, future uses might include three-dimensional imaging and collision avoidance, such as cameras detecting other vehicles in a car's blind spot. The camera also might become integrated into computer screens.

"This camera might follow the same life-cycle as the digital clock," Ackland says, "which used to be an expensive stand-alone device, but is virtually everywhere now."

Because the camera on a chip is an offshoot of today's silicon-chip technology, any semiconductor manufacturer could produce the chip at existing facilities, Ackland said. The same situation has not existed with CCD technology, which requires specially designed manufacturing facilities.

Lucent Technologies, headquartered in Murray Hill, N.J., designs, builds and delivers a wide range of public and private networks, communications systems and software, data networking systems, business telephone systems and microelectronic components. Bell Labs is the research and development arm of Lucent Technologies. For more information on Lucent Technologies, visit the company's web site at http://www.lucent.com.

Vanguard International Semiconductor Corporation (VIS), with headquarters in the Science-Based Industrial Park, Hsinchu, Taiwan, designs, develops, manufactures, markets and sells DRAM and other semiconductor products. VIS has its own in-house design expertise, supplemented by a strong US design subsidiary in San Jose, California. VIS sells its products through its US sales office, as well as a worldwide network of distributors and representatives. More information is available at http://www.vis.com.tw.

(Editor's note: Pictures of the CMOS "camera on a chip" are available at http://www.lucent.com/press/0798/index.html)