INTOXICATION DETECTION SYSTEM

Drug intoxication significantly affects a person’s behavior,
balance, and decision-making ability, creating serious risks
in public safety, transportation, and workplaces. Traditional
methods of detecting intoxication are manual, time�consuming, and often inaccurate. The objective of this
project is to develop an automated intoxication detection
system using facial image analysis. The proposed system
utilizes deep learning techniques, specifically a
MobileNetV2-based model, to classify individuals as
intoxicated or sober. The system involves dataset collection,
image preprocessing, feature extraction, and classification.
The processed image is analyzed by the trained model to
detect patterns associated with intoxication. The results are
displayed in a user-friendly interface, providing quick and
reliable predictions. Experimental results show that the
system achieves high accuracy and reduces human effort,
making it an efficient solution for real-time intoxication
detection