Implementasi Haar Cascade Untuk Deteksi Kendaraan Bermotor Pada Pemantauan Lalu Lintas Kota Malang

Rian Setya Budi, Deddy Rudhistiar, Nurlaily Vendyansyah

Abstract

Abstrak : 

Pemantauan lalu lintas di kota besar seperti Kota Malang semakin penting untuk mendukung ketertiban dan analisis jalan raya. Meskipun infrastruktur CCTV telah tersedia, teknologi deteksi kendaraan secara otomatis masih terbatas. Penelitian ini mengimplementasikan metode Haar Cascade untuk deteksi kendaraan bermotor pada rekaman CCTV guna meningkatkan akurasi dan efektivitas pemantauan lalu lintas. Evaluasi kinerja sistem dilakukan menggunakan metrik akurasi, presisi, recall, F1 score, serta pengujian blackbox untuk mengukur keandalan fungsi sistem. Hasil menunjukkan performa deteksi terbaik pada pagi hari, dengan akurasi 94.84%, presisi 84.21%, recall 100%, dan F1 score 91.43% untuk mobil, serta akurasi 93.81%, presisi 92.45%, recall 94.23%, dan F1 score 93.33% untuk motor. Namun, performa menurun pada malam hari, dengan akurasi 88.89% dan F1 score 71.43% untuk mobil, serta akurasi 46.15% dan F1 score 63.64% untuk motor. Pengujian blackbox menunjukkan bahwa setiap fungsi sistem berjalan sesuai dengan yang diharapkan tanpa ditemukan error dalam pengoperasian fitur utama. Penelitian ini menunjukkan bahwa metode Haar Cascade efektif untuk deteksi kendaraan di siang hari, namun membutuhkan peningkatan untuk kondisi pencahayaan rendah.

=================================================

Abstract :

Traffic monitoring in major cities like Malang has become increasingly important to support order and road analysis. Although CCTV infrastructure is available, the technology for automatic vehicle detection remains limited. This study implements the Haar Cascade method for motor vehicle detection on CCTV recordings to improve the accuracy and effectiveness of traffic monitoring. The system's performance evaluation was conducted using metrics such as accuracy, precision, recall, F1 score, and blackbox testing to assess system reliability. The results show the best detection performance in the morning, with an accuracy of 94.84%, precision of 84.21%, recall of 100%, and an F1 score of 91.43% for cars, and an accuracy of 93.81%, precision of 92.45%, recall of 94.23%, and an F1 score of 93.33% for motorcycles. However, performance decreased at night, with an accuracy of 88.89% and an F1 score of 71.43% for cars, and an accuracy of 46.15% and an F1 score of 63.64% for motorcycles. Blackbox testing shows that all system functions operate as expected without any errors in the main features. This study demonstrates that the Haar Cascade method is effective for vehicle detection during the day but requires improvement for low-light conditions.

Keywords

Haar Cascade; Deteksi Kendaraan; Evaluasi Kinerja; CCTV Lalu lintas; Kota Malang

Full Text:

PDF

References

[1] M. Harahap, J. Elfrida, P. Agusman, M. Rafael, R. Abram, and K. Andrianto, “Sistem Cerdas Pemantauan Arus Lalu Lintas Dengan YOLO (You Only Look Once v3),” Semin. Nas. APTIKOM, pp. 367–376, 2019.

[2] L. Chen, F. Ye, Y. Ruan, H. Fan, and Q. Chen, “An algorithm for highway vehicle detection based on convolutional neural network,” Eurasip J. Image Video Process., vol. 2018, no. 1, pp. 1–7, 2018, doi: 10.1186/s13640-018-0350-2.

[3] M. I. Ramadhani, A. E. Minarno, and E. B. Cahyono, “Vehicle Classification using Haar Cascade Classifier Method in Traffic Surveillance System,” Kinet. Game Technol. Inf. Syst. Comput. Network, Comput. Electron. Control, vol. 3, no. 1, pp. 57–64, 2017, doi: 10.22219/kinetik.v3i1.546.

[4] M. A. Kamil and Y. M. Djaksana, “JIIC : JURNAL INTELEK INSAN CENDIKIA PERBANDINGAN ALGORITMA CONVOLUTIONAL NEURAL NETWORK ( CNN ) DAN ALGORITMA YOU ONLY LOOK ONCE ( YOLO ) UNTUK DETEKSI WAJAH COMPARISON OF CONVOLUTIONAL NEURAL NETWORK ( CNN ) ALGORITHM AND YOU ONLY LOOK ONCE ( YOLO ) ALG,” no. November, 2024.

[5] A. Morchid et al., “IoT-enabled fire detection for sustainable agriculture: A real-time system using flask and embedded technologies,” Results Eng., vol. 23, no. July, p. 102705, 2024, doi: 10.1016/j.rineng.2024.102705.

[6] R. Romindo, O. P. Barus, and J. J. Pangaribuan, “Implementasi Metode Naive Bayes Classifier Terhadap Klasifikasi Topik Kemacetan Lalu Lintas Indonesia Melalui Tweet,” J. Media Inform. Budidarma, vol. 8, no. 2, p. 1087, 2024, doi: 10.30865/mib.v8i2.7470.

[7] P. Studi and T. Informatika, “Deteksi Perjalanan Kendaraan Untuk Mengukur Kepadatan Lalu Lintas Menggunakan Menggunakan OpenCV Python,” J. Ilmu Komput., vol. 3, no. 2, pp. 1–3, 2020.

[8] S. Elkerdawy, M. Elhelw, and R. Sayed, “Real-Time Vehicle Detection and Tracking Using Haar-Like Features and Compressive Tracking,” ResearchGate, vol. 252, no. November, pp. 381–390, 2013, doi: 10.1007/978-3-319-03413-3_27.

[9] P. Viola and M. Jones, “Robust Real-time Object Detection Paul,” Int. J. Comput. Vis., vol. 57, no. 2, pp. 137–154, 2004, doi: https://doi.org/10.1023/B:VISI.0000013087.49260.fb.

[10] Alexander Mordvintsev and Abid K, “OpenCV-Python Tutorials Documentation,” 2017. https://readthedocs.org/projects/opencv24-python-tutorials/downloads/pdf/stable/ (accessed Nov. 25, 2024).

[11] J. Zhang, X. Zou, L. D. Kuang, J. Wang, R. S. Sherratt, and X. Yu, “CCTSDB 2021: A More Comprehensive Traffic Sign Detection Benchmark,” Human-centric Comput. Inf. Sci., vol. 12, 2022, doi: 10.22967/HCIS.2022.12.023.

[12] A. Mashudi, F. Rofii, and M. Mukhsim, “Sistem Kamera Cerdas Untuk Deteksi Pelanggaran Marka Jalan,” JASEE J. Appl. Sci. Electr. Eng., vol. 1, no. 01, pp. 15–25, 2020, doi: 10.31328/jasee.v1i01.4.

[13] W. S. Pambudi and B. M. Simorangkir, “Facetracker menggunakan metode Haar Like Feature dan PID pada model simulasi,” J. Teknol. dan Inform., vol. 2, no. 2, pp. 142–154, 2012.

[14] Q. Wang, Q. Zhang, X. Liang, Y. Wang, C. Zhou, and V. I. Mikulovich, “Traffic lights detection and recognition method based on the improved yolov4 algorithm,” Sensors, vol. 22, no. 1, pp. 1–20, 2022, doi: 10.3390/s22010200.

[15] D. Indra, H. Herman, and F. S. Budi, “Implementasi Sistem Penghitung Kendaraan Otomatis Berbasis Computer Vision,” Komputika J. Sist. Komput., vol. 12, no. 1, pp. 53–62, 2023, doi: 10.34010/komputika.v12i1.9082.

[16] M. Hasanah et al., “Automatic Car Detection Using Haar Cascade Classifier and Convolutional Neural Network for Traffic Density Estimation,” Indones. J. Artif. Intell. Data Min., vol. 4, no. 1, pp. 11–18, 2021.

DOI: https://doi.org/10.20961/ijai.v9i2.95558

Refbacks

  • There are currently no refbacks.