Color Depth Mapper Imaging Sensor System
Autonomous and semi-autonomous robotic systems require information about their surroundings in order to navigate properly. A video camera machine vision system can supply position information of external objects, but no range information. Ideally, a system that, in one package, provides 3-dimensional relative information about external objects is needed. Under a NASA Langley Research Phase II Small Business Innovative Research Contract, Nanohmics is developing a lightweight, compact, low power, low cost, modular sensor system that produces a depth map of the surroundings. By combining a color optical camera, a multi-element range finding system, and digital processing electronics, a single low cost sensor system is being developed to provide relative position and anti-collision information. For additional reliability, the system will not include any moving parts. By making this sensor system modular, and with a simple-to-use serial interface, it can be used in various robotics applications including, but not limited to, autonomous and semi-autonomous planetary surface rovers. The figure demonstrates the proposed device geometry of the 3-Dimensional Vehicle Imaging Sensor for Incident Obstacle Navigation (3D VISION Mapper™). The proposed system could, for example, be mounted on the long-neck mast near the PANCAMs and NAVCAMs on Martian robotic rovers.
Existing laser range finding systems are expensive and consume large amounts of power. Additionally, they are sensitive to only a narrow solid angle, and must be scanned mechanically in order to provide more than a single dimension of depth information. The Nanohmics device under development provides depth information in two dimensions by means of a segmented optical detection system and photodiode array detectors. The imaging rangefinder system is essentially a “depth camera.” With a depth measurement precision of a few cm, and a range of about 25-50 m, a relatively low frequency of modulation is required, resulting in lower power consumption. Additionally, by reducing frame rates to 10 to 20 frames/sec, power and cost can be lowered further by multiplexing the photodiode output, thereby having only one set of expensive and power-hungry synchronous detection electronics.