New CMOS Chips See Colors

German researchers have produced a CMOS image sensing chip that can see colors that could help future smart cars recognize traffic signs, park the car, and warn their drivers of blind spot hazards. Driver assistance systems include high-tech cameras that must meet various requirements, such as being able to withstand high ambient temperatures, be small, light and robust and cost as little as possible.

Currently, CMOS sensors are used for these in-car systems. These semiconductor chips, which can translate light signals into electrical impulses, are used in most digital cameras. But the sensors used in industrial and other special cameras are typically color blind. So scientists at the Fraunhofer Institute for Microelectronic Circuits and Systems IMS in Duisburg have developed a new process for making CMOS image sensors that can see color. Usually the sensors are manufactured on silicon wafers using a semiconductor technique, known as the complementary metal oxide semiconductor (CMOS) process.

"We have integrated a color filter system in the process," said Prof. Dr. Holger Vogt of the IMS. "In the same way as the human eye needs color-specific cone types, color filters have to be inserted in front of the sensors so that they can distinguish color." The function is handled by polymers dyed in the primary colors red, green and blue. Every pixel on the sensor is coated with one of these three colors by a machine which covers the sensor disk propels with a micrometer-thick layer of polymer. Using UV light and a mask which is only transparent on the desired pixels, the dye is fixed at the required locations and the rest is then washed off.

Researchers have also created special microlenses which aid the sensor in more efficiently capturing and measuring the light. They create a separate lens with transparent polyimide for each individual pixel, almost doubling the light-sensitivity of the image sensor. Not only does the new CMOS process make it possible to cost-efficiently improve the performance of driver assistance systems, but endoscopes could also benefit from the properties of color-optimized CMOS image sensors.