Detection and identification of illegal-modified private cars through frequency band analysis
Post-purchase-illegal-modification of road vehicles' exhaust systems by vehicle fanatics for loudness has caused an upsurge in nuisances to local residences when driven and rallied during late night and early mornings. Traditional enforcement measures rely on setting up roadblocks by the Police at rallying hotspots. These measures involve the judgement of individual police officer and can be subjective. With the aid of noise monitoring equipment, the accuracy and efficiency of on-site exhaust system-modified vehicle detection can be enhanced. Since the sound profile of a single-vehicle exhaust cannot be captured by simple roadside noise level measurements alone, segregation and analysis of the noise spectrum are employed to identify vehicles with modified exhaust systems. The paper presents the findings of investigating the feasibility and accuracy of off-the-shelf devices for detecting vehicles fitted with modified exhaust systems, with private cars being the primary target. A pilot test by roadside noise monitoring has been conducted with a sound level meter and an acoustic camera and revealed an on-site accuracy of up to 75%. Data collected during the tests were further used to explore the applicability of integrating artificial intelligence with traditional noise monitoring devices.