Deteksi Objek menggunakan Dashboard Camera untuk Sistem Peringatan Pencegah Kecelakaan pada Mobil

Penulis

Raymond Sutjiadi, Timothy John Pattiasina

Abstrak

Saat ini penggunaan dashboard camera marak digunakan pada mobil untuk merekam kondisi sekitar kendaraan ketika berkendara. Dashboard camera adalah semacam kamera yang ditempatkan pada bagian dashboard mobil dengan kamera menyorot ke arah depan kendaraan yang berfungsi untuk merekam kondisi jalan. Di lain pihak, pada mobil premium saat ini sudah disematkan beberapa teknologi canggih untuk mencegah terjadinya kecelakaan atau tabrakan yang biasa disebut dengan Forward Collision Warning System. Teknologi ini pada dasarnya berfungsi untuk mencegah terjadinya tabrakan dari arah depan, baik dengan cara aktif ataupun pasif. Pada penelitian ini akan dibuat sebuah sistem terintegrasi dimana dashboard camera, yang diimplementasikan menggunakan kamera smartphone berbasis Android, tidak hanya digunakan untuk perekaman secara statis, tetapi juga digunakan untuk membuat sistem pencegah kecelakaan secara pasif. Adapun aplikasi ini dibuat dengan menggunakan metode pengolahan citra digital untuk mendeteksi keberadaan objek di depan mobil dengan menggunakan Tensorflow Open Source Machine Learning Library. Dari hasil pengujian tampak bahwa aplikasi ini mampu mendeteksi objek kendaraan berupa mobil penumpang, bus, dan truk, serta dapat memberikan peringatan baik secara visual maupun alarm apabila kendaraan di depan sudah berada pada jarak yang cukup dekat untuk memperingatkan pengemudi akan bahaya tabrakan.

 

Abstract

Nowadays dashboard camera becomes familiar to be used in a car to record the condition around the vehicle while driving. Dashboard camera is a video camera placed in car’s dashboard faces in front of the vehicle to record the road condition. In the other side, premium cars now equipped with advanced technology to prevent collision called Forward Collision Warning System. This technology basically acts to prevent front collision, either in active or passive ways. In this research was built integrated system where dashboard camera, which implemented by camera of Android based smartphone, not only used as static recording, but also as passive collision avoidance system. This application was developed using Object Detection Method in Tensorflow Open Source Machine Learning Library. The research stage was started from problem analysis, literature study to search comparison from previous research, also software development and finalized with testing to measure system performance. From the testing result, this application was able to detect vehicle objects in form of passenger car, bus, and truck, also could provide both visual and alarm warning when there was a vehicle come closely from in front, to warn the driver about the danger of collision.


Teks Lengkap:

PDF

Referensi


HNOOHOM, N. & THANAPATTHERAKUL, T., 2016. Image Problem Classification for Dashboard Cameras. IEEE Xplore: 2016 12th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS).

YADAV, G.K., KANCHARLA, T. & NAIR, S., 2011. Real Time Vehicle Detection for Rear and Forward Collision Warning Systems. ACC 2011: Advances in Computing and Communications, pp. 368-377.

ADLA, R., AL-HOLOU, N., MURAD, M. & BAZZI, Y.A., 2013. Automotive Collision Avoidance Methodologies: Sensor-Based and ITS-Based. ACS International Conference on Computer Systems and Applications (AICCSA), pp. 1-8.

CICCHINO, J.B., 2017. Effectiveness of Forward Collision Warning and Autonomous Emergency Braking Systems in Reducing Front-to-rear Crash Rates. Accident Analysis and Prevention, vol. 99, pp. 142–152.

DEVIKAR, P. 2016. Transfer Learning for Image Classification of Various Dog Breeds. International Journal of Advanced Research in Computer Engineering and Technology (IJARCET), Vol.5: 2707-2715.

HORN, B.K.P., FANG, Y. & MASAKI, I., 2009. Hierarchical Framework for Direct Gradient-based Time-to-contact Estimation. IEEE Intelligent Vehicle Symposium, pp. 1394-1400.

IHZAN, M.Z., WIDYAWAN & EDI, L.N., 2012. Deteksi Gerak Dengan Menggunakan Metode Frame Differences Pada IP Camera. Proceedings of International Conference on Information Technology and Electrical Engineering, pp. 52-55.

SUSMITA, A.M. & MALVIYA, A.V., 2013. Traffic Surveillance by Counting and Classification of Vehicles from Video Using Image Processing. International Journal of Advance Research in Computer Science and Management Studies, 1(6), 2321-7782.




DOI: http://dx.doi.org/10.25126/jtiik.2020712520