Smoke Detection Algorithm on Digital Image Sequences Using Gaussian Mixture Model (GMM) Segmentation Method and Region Growing for Fire Detection System
DOI:
https://doi.org/10.55749/ijdrr.v1i1.189Keywords:
CCTV, Detection Algorithm, GMM Segmentation, Image Segmentation, Region Growing, Smoke DetectionAbstract
Although various fire prevention efforts have been made, fire incidents continue to occur in various regions. To minimize the risk of these fires, a quick response is necessary so that mitigation efforts can be promptly carried out. This can be achieved by implementing a fire detection system. In such a system, smoke detection is one of the key elements aimed at providing a quick response to potential fire hazards. The development of a smoke detection algorithm is therefore crucial for enhancing the performance of the fire detection system. This study proposes a smoke detection algorithm that combines the Gaussian Mixture Model (GMM) segmentation method and region growing. The input data is in the form of video, which undergoes various preprocessing steps to prepare the data in the required format. Then, in the processing stage, two segmentation steps are carried out: GMM segmentation and region growing. GMM segmentation focuses on dividing the regions in the image into several clusters. Meanwhile, the region growing process is applied to expand the regions identified as smoke by considering the spatial context of the pixels. The output generated is a binary image with smoke objects in white and the background in black. Visually, the algorithm shows that the proposed approach is capable of providing better smoke detection compared to single segmentation methods. The results of this study demonstrate the potential of the smoke detection algorithm using GMM and region growing segmentation in improving the performance of fire detection systems.
References
[1] R. L. Schumann et al., “Wildfire recovery as a ‘Hot moment’ for creating fire-adapted communities,” International Journal of Disaster Risk Reduction, vol. 42, p. 101354, Jan. 2020, doi: 10.1016/j.ijdrr.2019.101354.
[2] P. Mulia, Nofrizal, and W. N. Dewi, “Analisis dampak kabut asap karhutla terhadap gangguan kesehatan fisik dan mental,” HEALTH CARE : JURNAL KESEHATAN, vol. 10, no. 1, 2021, doi: 10.36763/healthcare.v10i1.103.
[3] Hairani, “Implementasi prinsip state responsibility dalam pencemaran kabut asap lintas negara (Transboundary haze pollution) di Indonesia,” Universitas Syiah Kuala, 2019.
[4] A. Molina-Pico, D. Cuesta-Frau, A. Araujo, J. Alejandre, and A. Rozas, “Forest monitoring and wildland early fire detection by a hierarchical wireless sensor network,” J. Sens., vol. 2016, pp. 1–8, 2016, doi: 10.1155/2016/8325845.
[5] M. Ruslan, M. S. Al-Amin, and E. Emidiana, “Perancangan sistem fire alarm kebakaran pada gedung laboratorium XXX,” Jurnal Tekno, vol. 18, no. 2, pp. 51–61, Nov. 2021, doi: 10.33557/jtekno.v18i2.1412.
[6] Z. Liu, “Review of recent developments in fire detection technologies,” Journal of Fire Protection Engineering, vol. 13, no. 2, pp. 129–151, May 2003, doi: 10.1177/1042391503013002003.
[7] A. Bovik, “Handbook of image & video processing,” Academic Press, vol. 369, no. 1, 2013.
[8] L. Mengxin, W. Xu, K. Xu, J. Fan, and D. Hou, “Review of fire detection technologies based on video image,” 2013.
[9] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2016. doi: 10.1109/CVPR.2016.90.
[10] F. S. Najiha, “Deteksi asap rokok menggunakan segmentasi color Image processing,” 2016.
[11] M. Shrivastava and P. Matlani, “A smoke detection algorithm based on K-means segmentation,” in 2016 International Conference on Audio, Language and Image Processing (ICALIP), IEEE, Jul. 2016, pp. 301–305. doi: 10.1109/ICALIP.2016.7846590.
[12] H. Wang and Y. Chen, “A smoke image segmentation algorithm based on rough set and region growing,” J. For. Sci. (Prague)., vol. 65, no. 8, pp. 321–329, Aug. 2019, doi: 10.17221/34/2019-JFS.
[13] Arnita, F. Marpaung, F. Aulia, N. Suryani, and R. C. Nabila, Computer Vision dan Pengolahan Citra Digital. Pustaka Aksara, 2022.
[14] F. Dufaux, “Grand challenges in image processing,” Frontiers in Signal Processing, vol. 1, Apr. 2021, doi: 10.3389/frsip.2021.675547.
[15] R. Munir, Pengolahan Citra Digital dengan Pendekatan Algoritmik. Bandung: Informatika Bandung, 2004.
[16] G. Ramella and G. S. di Baja, “Color histogram-based image segmentation,” 2011, pp. 76–83. doi: 10.1007/978-3-642-23672-3_10.
[17] N. Iriawan, “Pemodelan mixture of mixture dalam pemilihan portofolio,” in Prosiding Seminar Nasional Statistika Universitas Diponegoro, 2011.
[18] G. J. McLachlan and D. Peel, Finite mixture models 2000. Wiley, 2000.
[19] T. M. Nguyen and Q. M. J. Wu, “Dirichlet Gaussian mixture model: Application to image segmentation,” Image Vis. Comput., vol. 29, no. 12, pp. 818–828, Nov. 2011, doi: 10.1016/j.imavis.2011.09.001.
[20] A. P. Dempster, N. M. Laird, and D. B. Rubin, “Maximum likelihood from incomplete data via the EM algorithm,” Journal of the Royal Statistical Society, vol. 39, no. 1, pp. 1–38, 1977, [Online]. Available: http://www.jstor.org/stable/2984875
[21] L. Handayani, “Identifikasi area kanker ovarium pada citra CT SCAN abdomen menggunakan metode Expectation Maximization,” Seminar Nasional Teknologi Informasi Komunikasi dan Industri (SNTIKI) 4, 2012.
[22] Y. Pu, J. Sun, N. Tang, and Z. Xu, “Deep expectation-maximization network for unsupervised image segmentation and clustering,” Image Vis. Comput., vol. 135, p. 104717, Jul. 2023, doi: 10.1016/j.imavis.2023.104717.
[23] H. Wang and Y. Chen, “A smoke image segmentation algorithm based on rough set and region growing,” J. For. Sci. (Prague)., vol. 65, no. 8, pp. 321–329, Aug. 2019, doi: 10.17221/34/2019-JFS.
[24] J. Qiao et al., “Data on MRI brain lesion segmentation using K-means and Gaussian Mixture Model-Expectation Maximization,” Data Brief, vol. 27, p. 104628, Dec. 2019, doi: 10.1016/j.dib.2019.104628.
