Design and characterization of a CMOS 3-D image sensor based on single photon avalanche diodes

The design and characterization of an imaging system is presented for depth information capture of arbitrary three-dimensional (3-D) objects. The core of the system is an array of 32 /spl times/ 32 rangefinding pixels that independently measure the time-of-flight of a ray of light as it is reflected back from the objects in a scene. A single cone of pulsed laser light illuminates the scene, thus no complex mechanical scanning or expensive optical equipment are needed. Millimetric depth accuracies can be reached thanks to the rangefinder's optical detectors that enable picosecond time discrimination. The detectors, based on a single photon avalanche diode operating in Geiger mode, utilize avalanche multiplication to enhance light detection. On-pixel high-speed electrical amplification can therefore be eliminated, thus greatly simplifying the array and potentially reducing its power dissipation. Optical power requirements on the light source can also be significantly relaxed, due to the array's sensitivity to single photon events. A number of standard performance measurements, conducted on the imager, are discussed in the paper. The 3-D imaging system was also tested on real 3-D subjects, including human facial models, demonstrating the suitability of the approach.

Introduction

VISION systems for capturing and rendering three-dimensional (3-D) images have existed since the dawn of photography. Despite considerable technological progress, especially in rendering techniques, 3-D vision has been relegated to research, particularly in robotics, and in relatively low-volume applications. Recently, the appearance of low-cost digital still and video cameras has reversed this trend and has accelerated the push to create low-cost and compact 3-D native vision systems. Currently, a number of novel and potentially high volume applications requiring fast and precise depth map evaluation are being devised and the market share of software for capturing, processing, and rendering 3-D images has grown exponentially. More recent applications include face recognition, virtual keyboards, object and person monitoring, land and sea surveying, virtual reality games, nonionizing medical tomographic imagers, stage and choreography analysis tools, etc. Cost concerns have prompted the emergence of a new generation of solid-state imagers that can achieve the desired accuracy and speed in compact implementations.