Conference proceedings

Ədəbiyyat 
1.
Viktor Mayer-Schönberger , Kenneth Cukier “BİG DATA”. Издательство «Манн, Иванов и Фербер» Москва, 2014 
 
 
PARKING FOR AUTONOMOUS VEHICLE USING DEEP LEARNING 
APPROACH 
Ebubekir CEYLAN 
1 *
, Fuad ALIEW
 
1
Gebze Technical University, Electronics Engineering, Turkey
0000-0002-8240-330X, [e-mail]
2 
Gebze Technical University, Electronics Engineering, Turkey 
0000-0002-0153-7868, [e-mail] 
* Responsible Author 
Abstract 
At the present time, thousands of traffic accidents occur every day due to human errors. These accidents cause human life or 
financial losses. Some of these accidents are accidents during the parking process of the vehicle. To solve this problem, we present an 
approach to parking for autonomous vehicles based on deep learning in order to detect an vacant parking lot and make a parking 
maneuver. It is aimed to perform these operations without any extra cost by using the camera in our vehicle, just like the cameras that 
are currently available in almost all vehicles. In this paper, we firstly designed an autonomous vehicle and present a deep neural network 
model based on convolutional neural networks, which gives very successful results in image recognition and classification tasks. 
Dataset which obtains in the test area and PKLot, CNRPark datasets, all of which are frequently used in this field, has used in order to 
train and evaluate the model. The autonomous vehicle takes images from the camera during the test and these images are entered into 
the model. The model prints out whether there is a parking space available for parking. According to this output, the vehicle either 
makes a parking maneuver or continues on its way to check the availability of the next parking space. This study showed that the 
vehicle we designed successfully parked without the need for human intervention. 
Keywords
: Autonomous Vehicle Design, Park Plot Detection, Convolutional Neural Network, Parking Maneuver, PKLot

---