Abbas El-Gamal, associate professor of electrical engineering at Stanford University, research associate Boyd Fowler, and graduate student David Yang are designing an advanced imaging sensor. They plan to integrate digitization with the image-capture process by moving it to the pixel level. "Pixel-level processing provides a number of potential advantages," says El-Gamal, including a dynamic range large enough to capture details of objects in bright sunlight and deep shade at the same time; reduction in noise; and pixel-level programmability, which could aid in automatic image recognition. Stanford has taken out four patents on different aspects of pixel-level processing. The device will be made from the same technology used to make low-power computer chips, CMOS. This allows engineers to combine the imaging sensors with computer circuitry, reducing the chip count and cutting production costs. CMOS imaging arrays are also faster than conventional CCD arrays because the pixels are read out in parallel while CCD arrays read out pixels sequentially, say the researchers. Other applications include digital imaging. Canon Inc., Eastman Kodak Co., Hewlett-Packard, Intel, and Interval Research made significant investments in the program. Industry partners will assist in design and prototyping, as well as fabrication of chips for test purposes. FAX: (415) 725-0247; e-mail [email protected].