Please use this identifier to cite or link to this item:
|Title:||Design and development of Image processing algorithms for quantitative road traffic data analysis||Authors:||Nair, Pranav Mohandas||Keywords:||DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision||Issue Date:||2014||Abstract:||The project aims to design and develop efficient algorithms using image processing techniques for the quantitative analysis of road traffic data in addition to enabling a better learning experience of a more commonly used software MATLAB. Road traffic data has been a major concern for traffic engineers in optimizing the efficiency and capacity of any modern transport system. The project aspires to develop a real time traffic analysis system for monitoring traffic flow, collect statistical data of traffic analysis and enhance the alogrithms in place to attain higher efficiency. Several existing traffic monitoring techniques such as edge detection, background difference, and inter-frame difference among others were researched and implemented in this project. Traffic video samples from express highways as well as city roads were collected for different lighting conditions, extracted into frames and subjected to different image processing techniques in MATLAB. Existing algorithms and Fuzzy Logic algorithms were implemented to obtain quantitative data such as vehicle speed and vehicle count and a comparative analysis was performed to obtain the better algorithm and better technique. Vehicle classification as per the size was incorporated in addition to measuring the percentage of road usage. The implementation and processing was done by designing a MATLAB Graphical User Interface by keeping in mind a myriad of possible user defined settings. Results comparison between the segmentation techniques showed that edge detection was the better method. In addition, a comparative study was done to observe which angle of video footage gave better results – the front angle or back angle. Also, comparison of the different lighting conditions was performed. A study of the results obtained when using different frame extraction rate, various window detection lengths was also done. In addition, a comparison of the results of the traffic in express highways and city roads was performed to observe if the results were aligned.||URI:||http://hdl.handle.net/10356/60974||Rights:||Nanyang Technological University||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.