Abstract:
This paper presents a detailed method for detecting whether a sailor is under the influence of alcohol by analyzing voice signals. The consumption of alcohol initiates a series of physiological changes in the body, beginning with its absorption into the bloodstream. As blood alcohol concentration rises, it stimulates increased blood circulation. One significant effect of alcohol is the engorgement of blood vessels in the nasal mucous membrane, which is the tissue lining the nasal cavities. This engorgement causes the blood vessels to expand, leading to nasal congestion. The resulting blockage can produce a distinct nasal quality in the speaker's voice, characterized by what is known as a widened formant bandwidth. This alteration in vocal quality can serve as an acoustic indicator of potential intoxication. Moreover, alcohol consumption has been found to increase lung capacity. This expansion allows for greater lung movement and alters the dynamics of air pulses produced during speech. As a consequence, the force exerted by the speaker's breath changes, which influences the way sound is produced. The spectrum of the sound may exhibit distinct characteristics, particularly in the lower and higher frequency ranges. Instead of showing sharp peaks typical of clear speech, the spectral representation tends to become flatter, indicating a loss of clarity in the vocal output. By analyzing these changes in voice signals—specifically, the nasal quality and the flattening of the sound spectrum—this method aims to establish a reliable indicator of alcohol influence in sailors.
Keywords:
Voice Signals, Speech Signal, Drinking Alcohol, Linear Predictive Coding
Citations:
APA:
Mun, L. S. (2022). A Critical Investigation into the Detection of Sailing Under the Influence Utilizing Linear Predictive Coding Techniques. Journal of Science and Engineering Management, 3(2), 11-26. https://doi.org/10.33832/jsem.2022.3.2.02