Single Frame Processing for Structured Light Based Obstacle Detection

We explore the capabilities of an inexpensive obstacle detection system consisting of a CCD color sensor, synchronously pulsed laser fan and supporting hardware and software. The novelty is the extreme simplicity of building such a system with commodity hardware, achieving a tiny form factor, low power operation and low cost. The camera, laser fan and supporting software constitute an active sensor that is mechanically passive, relying on the motion of its host platform to probe its surroundings. In this initial investigation, we determine useable range limits as a function of ambient lighting conditions by processing individual frames collected with a prototype system. We examine failure and recovery of the sensor when direct lighting sources come into the field of view. Future work will determine the efficacy of the full design embedded in various host platforms. Since the geometric relation between the optical sensor and each laser is fixed, we seek optimal parameters for this relationship given the relevant constraints of the chosen system. In situations with sufficient ambient light, we speed the computation of well known computer vision techniques for object identification to yield estimates of obstacle positions within the environment by incorporating range data obtained from the laser return. Synchronous pulsing of the laser with a short electronic shutter time on the optical sensor allows operation of the device as an ANSI Z136 class I device since the laser’s active duty cycle is highly compressed. This approach renders visible wavelengths invisible to the naked eye under most conditions.