Road Traffic Condition Estimation Based on Road Acoustics

Traffic density estimation is a big challenge in different cities in India for automatic traffic signaling system. Various techniques based on computer vision and image processing are proposed to address these issue. In some techniques costly sensors like magnetic loops, speed guns are also used. In this paper, a simple approach based on road acoustics is proposed to estimate the traffic density and to classify it into three categories as free, medium and jammed. The proposed technique is cost effective as it uses a very low cost omni-directional microphone. MFCC features are extracted from the acoustic signals and three different classifiers are used to classify the traffic density. The experimental evaluation shows that, MFCC features and vector quantization yields the best results for better classification.