Browsing by Author "Satar, Burak"
Now showing 1 - 2 of 2
- Results Per Page
- Sort Options
Item Deep learning based vehicle make and model classification(Uludağ Üniversitesi, 2018-09-11) Satar, Burak; Dirik, Ahmet Emir; Uludağ Üniversitesi/Fen Bilimleri Enstitüsü/Elektronik Mühendisliği Anabilim Dalı.Many pieces of research have been performed on the vehicle make & model classification recently. This thesis studies the problems regarding this topic. Being able to reach high classification accuracy is one of the main challenges as well as to reduce the annotation time of the images. In this thesis, it is first created a fine-grained dataset by using online marketplaces of Turkey to address these challenges by implementing all experiments on it. Then, it is proposed a pipeline to combine an SSD (Single Shot Multibox Detector) model with a CNN (Convolutional Neural Network) model. In the pipeline, the vehicles are detected by following an algorithm to diminish the time of annotation. The detected vehicles are fed into the CNN model. The results show that the classification accuracy reaches roundly 4% better score when compared with a conventional CNN model. Later, the detected vehicles are picked as Ground Truth Bounding Boxes (GTBB) of the images. Thus, every single image in the dataset contains its GTBB. As a result, they are fed into an SSD model in a different pipeline. By that, it is reached acceptable classification & detection accuracy results even though it is not used perfectly shaped GTBB. Lastly, it is proposed an application which focuses on a use case by using our proposed pipelines. Assuming that license plates are readable, it detects the unlawful vehicles by comparing their license plate numbers and make & models.Item Deep learning based vehicle make-model classification(Springer, 2018) Kurkova, V.; Manolopoulos, Y.; Hammer, B.; Iliadis, L.; Maglogiannis, I.; Satar, Burak; Dirik, Ahmet Emir; Uludağ Üniversitesi/Mühendislik Fakültesi/Elektrik-Elektronik Mühendisliği Bölümü.; Uludağ Üniversitesi/Mühendislik Fakültesi/Bilgisayar Mühendisliği Bölümü.; 0000-0002-6200-1717; K-6977-2012; 57204183877; 23033658100This paper studies the problem of vehicle make & model classification. Some of the main challenges are reaching high classification accuracy and reducing the annotation time of the images. To address these problems, we have created a fine-grained database using online vehicle marketplaces of Turkey. A pipeline is proposed to combine an SSD (Single Shot Multibox Detector) model with a CNN (Convolutional Neural Network) model to train on the database. In the pipeline, we first detect the vehicles by following an algorithm which reduces the time for annotation. Then, we feed them into the CNN model. It is reached approximately 4% better classification accuracy result than using a conventional CNN model. Next, we propose to use the detected vehicles as ground truth bounding box (GTBB) of the images and feed them into an SSD model in another pipeline. At this stage, it is reached reasonable classification accuracy result without using perfectly shaped GTBB. Lastly, an application is implemented in a use case by using our proposed pipelines which detects the unauthorized vehicles by comparing their license plate numbers and make & models. It is assumed that license plates are readable.