Smart Sound Sensor to Classify Infant Cries

Abstract

This thesis describes the process for implementing a smart sound sensor to categorize baby cries according to their fundamental frequencies. This is accomplished by integrating a sound sensor module with the digital signal processing software LABVIEW. An algorithm is created to determine the fundamental frequency after the sampled signal is conditioned to match the baby's cry's bandwidth. This program determines the phonation and hyperphonation of a cry by comparing the fundamental frequency values with benchmark values from literature reviews. LEDs actuated by an Arduino microprocessor are used to output the cry characteristics through light. The measurements also undergo statistical analysis.