| dc.description.abstract |
ABSTRACT
The concept of a smart city is the most important issue in the aspect of developing cities in the world. Where the city must promise a more comfortable, organized, healthy and efficient life. Smart transportation is one of the most important planning concepts for the formation of a smart city to improve the urban economy. With the existence of smart transportation, traffic information can be easily obtained by road users, including toll roads. The problem of congestion on toll roads is caused by users who have to stop and make toll road payments. Because some GTOs in Indonesia still have sensors that often fail to detect trailer trucks. To overcome this problem, a system was created to compare Pre-trained Alexnet and Mobilenetv2 to get the best accuracy in recognizing the types of cars or trucks that will enter the toll road using the Convolutional Neural Network (CNN) method. After testing, it was concluded that the given method was successful in identifying the type of car based on its shape and obtained an Alexnet accuracy of 92.71% and Mobilenetv2 93.98%.
Keyword : Automatic toll, Car classification, Deep learning, Convolutional Neural network (CNN), Alexnet, Mobilenetv2.
| en_US |
