Today, information technologies are
used in almost every stage of life. It seeks to find solutions too many issues
and problems. Image processing applications have been widely used in many areas
in recent years and are trying to solve problems. Many applications which
perform tasks such as classification, counting, measurement, target tracking
have been developed. The aim of this study is to provide a solution for
different applications using an effective and cost-effective method to detect
the brand and model of vehicles. A classification method is implemented using
deep neural network in the determination of the vehicle brand. The proposed
solution is tested on various images taken from different angles and obtained
from different sources. Faster-RCNN method which is one of deep neural networks
is used to brand detection of vehicles in this study. It is observed that
Faster-RCNN method performs 67.66% classification accuracy.
Vehicle brand detection Image Processing Deep Neural Networks Tensorflow Faster-RCNN
Birincil Dil | İngilizce |
---|---|
Bölüm | Research Article |
Yazarlar | |
Yayımlanma Tarihi | 30 Eylül 2019 |
Yayımlandığı Sayı | Yıl 2019 Cilt: 7 Sayı: 3 |
Address: Selcuk University, Faculty of Technology 42031 Selcuklu, Konya/TURKEY.