A MINIATURIZED SPACE-VARIANT ACTIVE VISION SYSTEM - CORTEX-I

We have developed a prototype for a miniaturized, active vision system with a sensor architecture based on a logarithmically structured, spa ce-variant, pixel geometry. The central part of the image has a high r esolution, and the periphery has a a smoothly falling resolution. The human visual system uses a similar image architecture. Our system inte grates a miniature CCD-based camera, a novel pan-tilt actuator/control ler, general purpose processors, a video-telephone modem and a display . Due to the ability of space-variant sensors to cover large work spac es, yet provide high acuity with an extremely small number of pixels, architectures with space-variant, active vision systems provide a pote ntial for reductions in system size and cost of several orders of magn itude. Cortex-I takes up less than a third of a cubic foot, including camera, actuators, control, computers, and power supply, and was built for a (one-off) parts cost of roughly US $2000. In this paper, we des cribe several applications that we have developed for Cortex-I such as tracking moving objects, visual attention, pattern recognition (licen se plate reading), and video-telephone communications (teleoperation). We report here on the design of the camera and optics (8 x 8 x 8 mm), a method to convert the uniform image to a space-variant image, and a new miniature pan-tilt actuator, the spherical pointing motor (SPM), (4 x 5 x 6 cm). Finally, we discuss applications for motion tracking a nd license plate reading. Potential application domains for systems of this type include vision systems for mobile robots and robot manipula tors, traffic monitoring systems, security and surveillance, telerobot ics, and consumer video communications. The long-range goal of this pr oject is to demonstrate that major new applications of robotics will b ecome feasible when small, low-cost, machine-vision systems can be mas s produced. We use the term ''commodity robotics'' to express the expe cted impact of the possibilities for opening up new application niches in robotics and machine vision, for what has until now been an expens ive, and therefore limited, technology.