Vehicle logo classification using bag of word descriptor and support vector machine classifier

• Main Author: Ng, Jia Phui
• Format: Thesis
• Language: English
• Published: 2019
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://eprints.utm.my/id/eprint/86087/1/NgJiaPhuiMSKE2019.pdf; http://eprints.utm.my/id/eprint/86087/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:132643

Intelligent Transportation Systems play an important role in traffic areas such as to record vehicular traffic data. In order to improve transportation safety and security, a system with the ability to automatically extract and recognize a vehicle is needed apart from the existing plate number recognition system. The detection and recognition of the vehicle type or model can be helpful in determining whether the vehicle is registered with the department of motor vehicle. Hence, this project aims at providing extra information with respect to the vehicle which is to determine the maker of the vehicles. In this project, the classification system is trained with 10 training images for each vehicle’s manufacturer. The common features for each logo will be extracted using the Speeded-Up Robust Features algorithm and then feature points will be grouped and arranged using Bag of Word representations which will then be clustered using K means clustering method. The vehicle’s classification will be determined by using Support Vector Machine classifier to classify and identify the logo of the vehicle. From the experimental results, the classification system achieved 87% and 77% for front view and side view images respectively with 1500, number of cluster.