A HIGH-RESOLUTION SENSOR BASED ON TRI-AURAL PERCEPTION

By virtue of their low cost and simplicity, ultrasonic sensors are widely used in time-of-flight ranging systems. Unfortunately, correctly interpreting the readings from such sensors proves to be very difficult. We present a high-resolution sensor composed of three ultrasonic sensors: one transmitter/receiver and two extra receivers, which allows a significant improvement in the information-extraction process. With this sensor we can determine the position, both distance and bearing, of all isolated objects in the field of view (almost-equal-to 25-degrees) using information contained in one single snapshot of a moderately complex scene. It is further shown that, within limits, the sensor system can also discriminate between different types of reflectors, based on their radius of curvature. In particular, the sensor can discriminate between walls and edges. These results are all based on the determination of the arrival times of the echoes present at the three receivers. In this respect, too, our sensor differs from the conventional ultrasonic sensor, which processes only the first echo to arrive at the receiver. A noise model, explaining the measured variations of the arrival times, is used to derive limits on the resolution of the results provided by the sensor. Furthermore, based on this model it is shown that, to a large extent, the results of the sensor are impervious to measurement variations common to all three receivers. Finally, this sensor is used in a realistic environment and the results are compared with those obtained from a conventional time-of-flight sensor.