An ultrasonic raytracing simulation method for bathroom surveillance

This paper presents a simulation method for bathroom surveillance using an ultrasonic raytracing technique. First, we explain the signal-capturing methods for the waves reflected by an object. Then, results obtained using this method are discussed. Object detection in an environment with significant interference is achieved through a signal-processing method, called beamforming. Evaluation results demonstrate that the proposed method achieved an average of 76.1% accuracy across seven different postures and positions, with an error range of 25 cm. Further evaluation shows that the results of our method were identical to the real measurements, which provides a foundation for future accuracy improvements through machine-learning techniques.