Journal ID : TRKU-14-05-2020-10747
[This article belongs to Volume - 62, Issue - 05]
Total View : 323

Title : Implementation of Illegal Parking Detection on Road Traffic using Neural Networks

Abstract :

Detection of objects that are not common is an important issue in the field of classification of vehicles to come. In object detection, many objects have been classified, such as cars, motorbikes, bicycles, and others. But there are still many objects that are not common, especially in Indonesia. There are still many objects that only exist in Indonesia but not in other countries. Like illegal parking, carts street vendors, and many more. Therefore the writer takes the object detection theme. This proposal proposes detecting unusual objects. The object that the author will detect is illegal parking. With the tools the authors will make, there will be a new classification of illegal parking. In this way a new classification of object detection will be born, namely illegal parking objects. So foreign people will not be confused by the situation they will encounter in Indonesia. The output of this research is the label of illegal parking classification and the probability value of the results of the illegal parking classification. The object detection system uses a faster R-CNN algorithm that can work well when detecting object distances and get accurate results for distances of 6 meters by 70%, 12 meters at 90%, 18 meters at 70%

Full article