DOI QR코드

DOI QR Code

데이터 증강 학습 이용한 딥러닝 기반 실시간 화재경보 시스템 구현

Implementation of a Deep Learning based Realtime Fire Alarm System using a Data Augmentation

  • 투고 : 2022.09.15
  • 심사 : 2022.09.25
  • 발행 : 2022.09.30

초록

본 논문에서는 딥러닝을 이용하여 실시간 화재경보 시스템을 구현하는 방법을 제안한다. 화재경보를 위한 딥러닝 학습 이미지 데이터셋은 인터넷을 통하여 1500장을 취득하였다. 일상적인 환경에서 취득된 다양한 이미지를 그대로 학습하게 되면 학습 정확도가 높지 않은 단점이 있다. 본 논문에서는 학습 정확도 향상을 위해 화재 이미지 데이터 확장 방법을 제안한다. 데이터증강 방법은 밝기 조절, 블러링, 불꽃사진 합성을 이용해 학습 데이터 600장을 추가해 총 2100장을 학습했다. 불꽃 이미지 합성방법을 이용하여 확장된 데이터는 정확도 향상에 큰 영향을 주었다. 실시간 화재탐지 시스템은 영상 데이터에 딥러닝을 적용하여 화재를 탐지하고 사용자에게 알림을 전송하는 시스템이다. Edge AI시스템에 적합한 YOLO V4 TINY 모델을 custom 학습한 모델을 이용해 실시간으로 영상을 분석해 화재를 탐지하고 그 결과를 사용자에게 알리는 웹을 개발하였다. 제안한 데이터를 사용하였을 때 기존 방법에 비하여 약 10%의 정확도 향상을 얻을 수 있다.

In this paper, we propose a method to implement a real-time fire alarm system using deep learning. The deep learning image dataset for fire alarms acquired 1,500 sheets through the Internet. If various images acquired in a daily environment are learned as they are, there is a disadvantage that the learning accuracy is not high. In this paper, we propose a fire image data expansion method to improve learning accuracy. The data augmentation method learned a total of 2,100 sheets by adding 600 pieces of learning data using brightness control, blurring, and flame photo synthesis. The expanded data using the flame image synthesis method had a great influence on the accuracy improvement. A real-time fire detection system is a system that detects fires by applying deep learning to image data and transmits notifications to users. An app was developed to detect fires by analyzing images in real time using a model custom-learned from the YOLO V4 TINY model suitable for the Edge AI system and to inform users of the results. Approximately 10% accuracy improvement can be obtained compared to conventional methods when using the proposed data.

키워드

과제정보

This work was supported by Seokyeong University in 2021

참고문헌

  1. Jong-Chan Park, Dae-Seong Kang, "Real-time video fire detection based on YOLO in antifire aurveillance systems," Proceedings of KIIT conference, KIIT, pp.179-181, 2021. DOI: 10.1007/s11554-020-01044-0
  2. Geun-Su Kim, Gyu-Do Park, and Soo-Hyeok Kang, "Fire Identification Embedded System Using YOLO algorithm," Proceedings of IEEK conference, pp.1920-1923, 2021.
  3. Young-Jin Kim, Eun-Gyung Kim, "Image based Fire Detection using Convolutional Neural Network," Journal of the Korea Institute of Information and Communication Engineering, Vol.20, No.9, pp. 1649-1656, 2016. DOI: 10.6109/jkiice.2016.20.9.1649
  4. J. Redmon, et. al., "You Only Look Once: Unified, Real-Time Object Detection," in Proc. CVPR, 2016. DOI: 10.48550/arXiv.1506.02640
  5. W. Liu, et. al., "SSD: Single Shot MultiBox Detector," in Proc. ECCV, 2016.
  6. T.-Y. Lim, et. al., "Focal Loss for Dense Object Detection," in Proc. ICCV, 2017.
  7. R. Girshick, et. al., "Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmenation," in Proc. CVPR, 2014.
  8. R. Girshick, "Fast R-CNN," in Proc. ICCV, 2015. DOI: 10.1109/CVPR.2014.81
  9. S. Ren, "Faster R-CNN: Toward Real-Time Object Detection with Region Proposal Networks," in Proc. NeurIPS, 2015.
  10. Jun-Mock Lee, Dae-Sung Kang, "Optimizing data augmentation based few-shot learning method for improving indoor fire detection accuracy," in Proc. Korea Digital Contents Societry, pp.237-240, 2019.
  11. I. Goodfellow, J. Pouget-Abadie and M. Mirza, "Generative Adversarial Networks," Neural Information Processing Systems, Vol.2, pp.2672-2680, 2014. DOI: 10.1109/MSP.2017.2765202
  12. You-min Na, Dong-hwan Hyun, Do-hyun Park, Se-hyun Hwang, Soo-hong Lee, "AI Fire Detection & Notification System," Journal of the Korea Society of Computer and Information, Vol.25, No.12, pp63-71, 2020. DOI: 10.9708/jksci.2020.25.12.063
  13. Cubuk, E. D., Zoph, B., Mane, D., Vasudevan, V., & Le, Q. V., "Autoaugment: Learning augmentation policies from data," arXiv preprint arXiv:1805.09501., 2018.
  14. M. Tan, et. al., "EfficientDet: Scalable and Efficient Object Detection," in Proc. CVPR, 2020.
  15. Bochkovskiy, Alexey, Chien-Yao Wang, and Hong-Yuan Mark Liao, "YOLOv4: Optimal Speed and Accuracy of Object Detection," arXiv preprint arXiv:2004.10934, 2020.
  16. Jae-Jung Kim, Jin-Kyu Ryu, Dong-Kurl Kwak, Sun-Joon Byun, "A Study on Flame Detection using Faster R-CNN and Image Augmentation Techniques," j.inst.Korean.electr. electron.eng., Vol.22, No.4, pp.1079-1087, 2018. http://dx.doi.org/10.7471/ikeee. 2018.22.4. 1079.
  17. Jonathan Hui, "mAP (mean Average Precision) for Object Detection,"https://jonathan-hui.medium.com/map-mean-average-precision-for-objectdetection-45c121a31173