SISTEM PENDETEKSIAN KECEPATAN KENDARAAN MENGGUNAKAN HAAR CASCADE BERBASIS RASPBERRY PI
DOI:
https://doi.org/10.21009/03.SNF2022.01.FA.18Abstract
Abstrak
Penelitian ini bertujuan mengembangkan sistem yang dapat mengidentifikasi kendaraan di jalan raya dan mendeteksi kecepatan kendaraan-kendaraan tersebut. Sistem ini dikembangkan pada Raspberry Pi menggunakan Python 3.7. Metode pada penelitian ini menggunakan metode Haar cascade untuk melakukan pengidentifikasian dan pelacakan obyek kendaraan pada rekaman video sehingga nilai kecepatan kendaraan dapat diestimasi. Pengujian dilakukan dengan menggunakan rekaman-rekaman video kendaraan di jalan raya dengan variasi nilai kecepatan yang berbeda. Dua metode analisis digunakan untuk mengevaluasi performa sistem dalam mengidentifikasi kendaraan dan mendeteksi kecepatannya. Nilai kecepatan asli dan jumlah kendaraan pada rekaman video dibandingkan dengan nilai yang dideteksi untuk mengukur performa sistem. Pengidentifikasian kendaraan dengan metode Haar cascade memperoleh nilai precision sebesar 96,32% dan nilai recall sebesar 96,91%. Pendeteksian kecepatan kendaraan yang dilakukan dengan program ini memperoleh nilai rata-rata error sebesar 3,37% dengan rata-rata rentang sebesar ±1,72 km/jam.
Kata-kata kunci: pendeteksian citra, pendeteksian kecepatan kendaraan, Haar cascade, Raspberry Pi.
Abstract
This study aimed to develop a system to identify vehicles on the highway and detect their speed. The system is developed on Raspberry Pi 2 using Python 3.7. In this study, Haar cascade is used to identify and track the vehicle objects on video records so that the speed can be measured. The program was tested using videos that show vehicle movement on the highway with various speed values. Two analysis methods were used to evaluate the performance of the system in identifying vehicles and detect their speed. The actual vehicle speed and the actual amount of vehicles on the video records were compared with the detected values to measure the system’s performance. Vehicle identification using Haar cascade on this research obtains 96,32% precision and 96,91 recall. Vehicle speed detection using this program obtains 3,37% average error with interval of ±1,72 km/hours.
Keywords: image detection, vehicle speed detction, Haar cascade, Raspberry Pi.
References
[2] M. E. Goodon, “Technical Shortcomings of Doppler Traffic Radar,” in Journal of Forensic Sciences, Lewisville, TX, pp. 1186-1193, 1985.
[3] F. C. Febrianto & F. Utaminingrum, “Perhitungan Kecepatan Secara Otomatis Menggunakan Metode Frame Difference Berbasis Raspberry Pi,” in Jurnal Fakultas Ilmu Komputer, state, pp. 10968-10974, 2019.
[4] S. S. Mohamed, N. M. Tahir & R. Adnan, “Background Modelling and Background Substraction Performance for Objecte Detection,” in 6th International Colloqium on Signal Processing & Its Applications, Melaka, pp. 236-241, 2010.
[5] Indrabayu et al., “Vehicle Detection and Tracking using Gaussian Mixture Model and Kalman Filter,” in 2016 International Conference on Computational Intelligence and Cybernetics, Makassar, pp. 115-119, 2016.
[6] P. Viola & M. Jones, “Rapid Object Detection using a Boosted Cascade of Simple Features,” in 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Cambridge, pp. 1-9, 2001.
[7] H. Schneiderman & T. Kanade, “A Statistical Method for 3D Object Detection Applied to Faces and Cars,” in Proceedings IEEE Conference on Computer Vision and Pattern Recognition, Hilton Head SC, pp. 746-751, 2000.
[8] H. A. Rowley, S. Baluja & T. Kanade, “Neural Network-based Face Detection,” in IEEE Transactions on Pattern Analysis and Machine Intelligence, Toronto, pp. 22-38, 1998.
