Vehicle Logo Recognition using RCNN for Intelligent Transportation Systems

Vehicle logo recognition systems has a significant role in applications such as traffic monitoring, security system and surveillance system. To categorize variant vehicle logos with higher accuracy, deep learning based RCNN (Recurrent Convolutional Neural Network) is proposed. The quantitative analysis of the proposed RCNN are related with the state of art techniques viz., SVM, k-NN, Naive Bayes and Decision tree with HOG features in terms of recognition parameters like accuracy, specificity, sensitivity and precision values.

I. Introduction

Pattern recognition is one of the important branch of digital image processing. It makes the computers to interpret the patterns in similar way of how the human does and in some cases more efficient than humans. In pattern recognition, logo recognition is a vital part which has a key part of image processing. The method of interpreting and identifying the logos automatically without human intervention is coined as logo recognition. It works efficiently by avoiding human mistakes which needs patience and accuracy. Logos are the design or patterns composed of characters and thoughtful drawings and it has peculiar features viz.,symbol, color and shape. Logos are widely used in variant fields with unique logos for each companies for example government and non-government organizations, educational organizations, private companies transportation systems etc. Intelligent transportation system is one of the domains that uses logo recognition and applied in current technology for monitoring and controlling the transportation systems.