Metode Star Skeletonization untuk Menghitung Jumlah Pejalan Kaki pada Citra
DOI:
https://doi.org/10.61769/telematika.v11i1.138Keywords:
Star Skeletonization, background subtraction, opening, closing, ekstraksi border.Abstract
Menghitung pejalan kaki secara otomatis, dilakukan dengan mendeteksi dan menganalisis objek-objek yang terlihat bergerak. Masalah yang ada jika dua orang atau lebih berjalan berdekatan dan menempel, orang dalam keadaan diam ditempat atau duduk, dan ketika objek yang bergerak bukan saja pejalan kaki tetapi orang naik sepeda, mobil, atau objek bukan pejalan kaki lainnya. Metode Star Skeletonization diterapkan untuk mengenali objek pejalan kaki berdasarkan 3 dan 5 garis yang ditentukan. Pra pemrosesan yang dilakukan background subtraction untuk mendapatkan objek bergerak, opening untuk menghilangkan noise, closing untuk menghilangkan celah pada objek, dan ekstraksi border untuk mendapatkan segmentasi objek. Hasil pengujian nilai threshold 5, structuring elemen berbentuk rect 3x3 untuk proses opening, structuring elemen berbentuk rect 5x5 untuk proses closing, dan Star Skeletonization tiga garis memberikan hasil terbaik untuk pengenalan pejalan kaki. Hasil pengujian untuk rentetan citra diam dengan pejalan kaki tunggal dan tidak berdempetan tingkat akurasi 90%, artinya objek pejalan kaki dapat dikenali dengan baik. Ketika ramai pejalan kaki tingkat akurasi 32.5%, hal ini disebabkan banyaknya pejalan kaki yang berdekatan sehingga nampak saling bergabung menjadi satu objek. Hal itu menyebabkan kesalahan perhitungan. Pra pemrosesan awal tidak dapat memisahkan objek yang tumpang tindih.
Count up of pedestrians are automatically, performed by detecting and analyzing objects appear to move. The problem occurs when two or more runs adjacent and attached, people who do not move or sit, and when a moving object is not only pedestrians but people ride bicycles, cars, or other objects rather pedestrian. Star Skeletonization methods applied to recognize objects pedestrians by 3 and 5 lines are determined. Initial processing done that background subtraction to get moving objects, opening to remove noise, closing to eliminate the gap on the object, and the border extraction to get the object segmentation. Results of testing the threshold value 5, 3x3 rect-shaped structuring element for the process of opening, 5x5 rect-shaped structuring element for the process of closing, and Star Skeletonization three lines provide the best results for the introduction of a pedestrian. The test results for a series of still image with a single pedestrian and not huddled get a 90% accuracy rate, meaning that the object of pedestrians can be recognized well. When crowded pedestrian accuracy rate of 32.5%, this is due to the many pedestrians adjacent to each other seem to merge into one object. That caused a calculation error. Initial processing can not separate the objects that overlap.
References
Fujiyoshi, Hironobu, Alan J. Lipton, dan Takeo Kanade. Real-time human motion analysis by image skeletonization. IEICE Transactions on Information and Systems 87.1 (2004): 113-120.
Gould, Stephen, Tianshi Gao, dan Daphne Koller. 2009. Region-based segmentation and object detection. Advances in neural information processing systems (2009): 655-663.
Irawan, E.W. 2012. Aplikasi Keamanan Pendeteksian Gerak Dengan Menggunakan Background Subtraction Dan Thresholded Euclidean Difference. Institut Teknologi Harapan Bangsa.
Lefloch, Damien, et al. 2008. Real-time people counting system using a single video camera. Electronic Imaging 2008. International Society for Optics and Photonics, 2008.
Merad, Djamel, K-E. Aziz, dan Nicolas Thome. 2010. Fast people counting using head detection from skeleton graph. In Advanced Video and Signal Based Surveillance, Seventh IEEE International Conference on (2010): 233-240.
__________. 2015. OPENCV (Open Source Computer Vision). http://opencv.org (diakses pada tanggal 16 Februari 2015).
Rakesh, Gupta, dan Kaur Rajpreet. 2008. Skeletonization algorithm for numeral patterns." International Journal of Signal Processing, Image Processing and Pattern Recognition 1.1 (2008): 63-72.
Chan, Antoni. 2015. VISAL (Video, Image, and Sound Analysis Lab). http://visal.cs. cityu.edu.hk (diakses pada tanggal 22 Januari 2015).
Wang, Xiaogang. 2007. Unsupervised Activity Perception by Hierarchical Bayesian Model. http://www.ee.cuhk.edu.hk/~xgwang/ HBM.html (diakses pada tanggal 2 April 2015).
Downloads
Published
Issue
Section
License
You are free to:
- Share — copy and redistribute the material in any medium or format for any purpose, even commercially.
- Adapt — remix, transform, and build upon the material for any purpose, even commercially.
- The licensor cannot revoke these freedoms as long as you follow the license terms.
Under the following terms:
- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.
- No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.
Notices:
You do not have to comply with the license for elements of the material in the public domain or where your use is permitted by an applicable exception or limitation.
No warranties are given. The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material.
