Investigation of noise features of different types of vehicles using an array of low-cost web cameras.
Hasan, Md Mehedi.
Date of Issue2013
School of Electrical and Electronic Engineering
Sound of moving vehicle provides important information of noise features of particular vehicles. In this final year project report, I will introduce using an array of low cost web/smart phone/video camera to finding vehicles noise and classification of vehicles by extracted raw data from vehicles noise using digital signal processing concept, originally used in human hearing reorganization (20Hz-20 kHz) to model the sound frequency distribution. In this paper will show that, this method can be very reliable to find vehicle sound classification if raw data are properly collect from moving vehicle and categorized them. I treat sampling frequency 48 kHz to reduce any aliasing or unwanted noise since both video camera and smart-phone camera frequency response range 20Hz to 20 kHz. In this frequency range, author analysis frequency spectrum for each type of vehicles (Motorcar, Lorry, and Motorbike) in time domain and frequency domain. A collection of vehicle sound sampled (real time traffic flow) use for further data analysis for data record. The peak of frequency spectrum is most important part of this experiment, where vehicles sound shows that audio/video capturing by video camera motorbike and lorry produces maximum level of noise and car is the lowest. Conversely, vehicle sound spectrum shows that audio/video capturing by smart phone camera work at low frequency (0-1000Hz), where lorry and motorcar produces maximum noise and motorbike at average level. After comparison among different set of data to conclude different types of vehicle produce different noise features. Author hope this experiment could be very useful to understand noise features of vehicles and can be use as a prototype for future research in vehicle noise investigation.
DRNTU::Engineering::Electrical and electronic engineering
Final Year Project (FYP)
Nanyang Technological University