Visual detection of vehicles using a bag-of-features approach

• Author(s): Pinto, Pedro ; Tomé, Ana Maria ; Santos, Vitor
• Date: 2013
• Persistent ID: http://hdl.handle.net/10773/10910
• Origin: RIA - Repositório Institucional da Universidade de Aveiro
• Subject(s): Intelligent vehicles; Object recognition; Machine learning algorithms; Support vector machines; Genetic algorithms
• Document Type: Conference Object
• Language: English

Description

This paper presents and evaluates the performance of a method for vehicle detection using a bag-of-features methodology. The algorithm combines Speeded Up Robust Features with a Support Vector Machine. An optimization to the bag-of-features dictionary based on a genetic algorithm for attribute selection is also described. The results obtained show that this method can successfully address the problem of vehicle classification.