Signal Noise Reduction Algorithm for Object Localization and Tracking Using a 2-Dimensional Ultrasonic Sensor Array

Ultrasonic sensors return time of flight (TOF) readings; by producing inaudible sound waves and calculating the time they take to return to the sensors. When arranged in an array (containing multiple sensors) they can have various applications, both independently and in conjunction with other transducers and cameras. One of the most common applications that has been developed in the last couple of decades has been object detection, localization and tracking. This research paper presents a novel algorithm that addresses one of the biggest challenges with using ultrasonic sensor arrays for object localization - signal noise. The proposed system also includes a laser module mounted on a pan-tilt motor mechanism to track objects, utilizing the signal noise reduction and angles calculation algorithm. The improvement in signal quality has been quantified with signal-to-noise ratio (SNR) and mean squared error (MSE). The presented algorithm will significantly improve the application of ultrasonic sensors, in autonomous vehicles, robotics and industrial automation.