Forest Fire Detection using Deep Learning

Abstract

The threat posed by forest fires to humanity has increased. In addition to giving a lot of living things a place to live and protection, they have also been a significant source of food, wood, and a lot of other goods. Throughout the dawn of time, woods have been integral to social, economic, and religious endeavors and have improved human life in numerous tangible and intangible ways. We must exercise sufficient caution when making decisions that could ultimately result in a tragic outcome if we want to safeguard our environment from these forest fires that are spreading quickly. Thus, we suggest an image identification technique based on convolutional neural networks for the early detection of forest fires (CNN).

Introduction

I. INTRODUCTION

It is important to recognise the value of forests. The air we breathe and the wood we use are both heavily dependent on trees for our life. In addition to offering homes for animals and people, forests also safeguard river basins, stop soil erosion, and slow down climate change. The fire detection method based on the computer vision system reached its zenith and replaced many conventional approaches. It is now playing a crucial part in our day-to-day lives as the world moves forward in the technological era. The backwoods is an enormous surface and the region is loaded up with trees, heaps of dried leaves, wood, etc. These components empower the fire when it begins. Fire can be touched off through many reasons like high temperatures in summer seasons, smoking, or a few gatherings which have firecrackers. When a fire begins, it will stay until it is totally recognized. The harm and the cost for recognizing fire in light of a timberland fire can be diminished when the fire is distinguished right on time as could be expected. Thus, fire identification is significant in this situation. Tracking down the specific area of the fire and sending a notice to the fire specialists not long after the event of the fire can have a constructive outcome. There are various kinds of fire discovery strategies utilized by the Public authority specialists, for example, satellite checking, tower observing, utilizing sensors, optical cameras, etc.

 II. LITERATURE SURVEY

Fire can be identified by utilizing how much smoke. The smoke sensors are utilized to gauge how much smoke from the fire, and it very well may be contrasted and an edge esteem and assuming it is past that worth, it is considered as a fire scenario.Using picture handling, fire can be distinguished straightaway. Fixing the CCTV camera all over and the pictures from these cameras can be handled to screen the fire. In the event that any progressions happen, it is not difficult to rapidly identify and douse the fire. This framework has a water dowser to smother the fire when the caution turns on. The CCTV camera is utilized for recording the video of a specific spot and it is associated with a smaller than expected PC called Raspberry-pi. So it could get the steady video recording of a specific region. The caught video pictures are handled edge by outline and when the fire is identified, the alert would be turned on. Likewise, the alert would be switched off when the fire stifled completely.The Virtual Organization Registering is utilized for the execution of the program, where the subtleties of video are moved from the raspberry-pi to the survey computer.This framework incorporates discovery, alert, fire smother, programming and organization modules. The calculation is delicate to enormous woodland regions and the location of fire in less probabilistic fire-inclined regions is awful. A Quicker R-CNN has been proposed to identify wood fires. To settle the equivocalness in preparing information non-genuine smoke pictures were delivered by setting up genuine or simulative smoke into the woods foundation. Be that as it may, this location calculation is more inclined to simulative wood smoke, and subsequently it requires further upgrades in supporting the exhibition with constant fire recognition. Out of control fire's smoke recognition strategy has been created to give a quick, cheap, and dependable framework for checking fire-delicate regions.

Almeida, T., & Pereira, J. M. [1] presented a deep learning-based approach using convolutional neural networks (CNNs) for forest fire detection using satellite imagery. The proposed method achieved high accuracy in detecting forest fires.

Kavak, S. [2] proposed a forest fire detection system based on deep learning techniques, specifically using a deep convolutional neural network (DCNN). The model demonstrated promising results in accurately identifying fire regions.

Zaharescu, A., Le Saux, B., & Datcu, M. [3] designed a paper that focuses on detecting forest fires from satellite images using CNNs. The authors propose a multi-scale CNN architecture to capture fire-related features at different levels of abstraction, leading to improved accuracy.

Mishra, S., Han, J., & Bhattacharya, M. [4] presented a deep learning-based forest fire detection system that utilizes a CNN to analyze video frames captured by surveillance cameras. The proposed system achieved high accuracy and real-time performance.

Azadbakht, M., & Saadatseresht, M. [5] as the authors proposed a model using deep learning algorithms based on VGGNet and ResNet architectures to detect forest fires. The models were trained and evaluated using high-resolution satellite imagery, and promising results were obtained.

Ren, L., Song, Z., & Ding, L. [6] proposed a paper that explores the use of deep learning for forest fire detection using remote sensing data. The authors propose a model based on the fusion of convolutional and long short-term memory (LSTM) networks to capture both spatial and temporal information, enhancing detection accuracy.

III. METHODOLOGY

Backwoods Fire is one of the catastrophes which is obliterating an exceptionally enormous area of prolific land. Numerous philosophies are being utilized to recognize wood fires as quickly as time permits. These techniques can be named video observing, WSN, what's more, fire discovery utilizing satellite pictures.

These cameras utilize infrared spectrometers and warm envisioning that can distinguish heat current but since of environmental circumstances, for example, haze, the shadow it can lead to a high deception rate. Satellite imaging can likewise be utilized in early location of fire flare-up. In this framework a base station investigates the information sent by satellite and recognizes areas of interest yet the framework execution is corrupted by the presence of mists.

Remote Sensor Organization can likewise be utilized for early location of a backwoods fire. In this there is an organization of different sensors that are utilized for fire discovery. In the proposed approach we are utilizing GSM and GPS along with sensors. This model will send the information to three distinct stations having the specific area of the flare-up.

Using the SQL database, we store the images of forests and fire and a combination of both in order to detect and predict forest fires.

IV. PROPOSED SYSTEM

Our motto is to predict and detect if there is any alarming alert of fire in the forest such that we can alert the forest officials and activate the water sprinkler.

The proposed approach utilizes calculations to remove highlights in light of picture handling methods and arrangements in view of PC supported knowledge techniques. With regards to existence, our primary step is to subsample the picture groupings, then, at that point, identify the development zone, which removes the smoking zones being referred to.

The applicant locales are then handled utilizing a smoke variety investigation calculation, a sharp edge finder, furthermore, discovery strategies for developing and expanding regions. Brain classifiers are then used to gauge the regions that portray the smoking regions in various settings.

It was a viable methodology that functioned admirably considerably under not exactly optimal circumstances. Unfortunate perceivability conditions can prompt bogus up-sides yet don't increment deceptions. An ongoing smoke location has been proposed for the procedures of collection of forefront pictures and optical stream.

This procedure was utilized to extricate the development properties of smoke and flares and to recognize them by their variety of unsettling influences. Their strategy is more probable to stifle commotion and constrict it totally in progressive edges. Limitations were restricted to light-delicate articles. For instance, lights turned on and off, and so on. In any case, their method knew how to recognize the other fire-like articles.

V. IMPLEMENTATION AND RESULT

Data is in the form of video frames which are obtained from CCTV footage, but for ease custom made videos are to be used to perform training and test. The collection of such videos with fire is a tedious task. The frames with fire and without fire are then stored as respectively.

Then we divide the dataset as a training set and a test set. This is to be done with great care because if the data fed to the neural network is faulty, the results will be corrupted and fail to produce an accurate system.

For the neural network to accurately detect fire, it needs to know the features of fire, how it looks in the computer's vision. The feature of fire is easily identifiable by the human eye. Fire emits reddish color; it has a shape under different circumstances and motion depending on the fuel it uses to burn. In this paper, the shape, color and motion of fire and smoke is used for the detection. We extract the features from different frames in the training set. The neural network extracts these features using the feature extraction network in the CNN which is powered by a custom algorithm. After extracting the features these video frames are classified into fire and non-fire scenarios. The features are extracted using bounding boxes using image descriptors.

The extracted features are then passed to the network to build a model. This model is a set of thresholds to help the network to accurately detect fire. The model learns from the features extracted and sets a standard for analyzing new input data. Validation of the machine learning model is essential because it is clearly important to get the accuracy and see if the system is working. The validation process is executed using another set of video frames which is completely unique from the dataset provided to build the model. According, the test results the system achieved about 93 % accuracy with the validation set.

Conclusion

The findings of the project are greatly satisfying. The system detected fire with an accuracy rate of 93 %. The results obtained show promise for implementation of Convolutional neural networks for detecting fire compared to other neural networks. The system combines several training data intelligently for calculating and reducing false alarm rates with fully connected networks. Then this data is passed to a decision-making algorithm to classify whether there is a fire or not. Although it has minor detection errors in some images, the overall performance and statistics are super-efficient. The only downfall is that it is a bit slow because it needs more computational power to produce results. The score of false alarm may be reduced by cleaning the data more and more. When implementing the rate of false alarm should be kept to minimum. The scope of using video frames in the detection of fire using machine learning is challenging as well as innovative. If this system with less error rate can be implemented at a large scale like in big factories, houses, forests, it is possible to prevent damage and loss due to random fire accidents by making use of the Surveillance systems. The proposed system can be developed to a more advanced system by integrating wireless sensors with CCTV for added protection and precision. The algorithm shows great promise in adapting to various environments.

References

[1] E. Menaka, S. S. Kumar and P. Parameshwari, \"Assessment of sparse forest and fire detection using threshold watershed algorithm,\" IET Chennai 3rd International on Sustainable Energy and Intelligent Systems (SEISCON 2012), Tiruchengode, 2012, pp. 1-6. doi: 10.1049/cp.2012.2196 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6719102&isnumber=6719080 [2] M. Gure, M. E. Ozel, H. H. Yildirim and M. Ozdemir, \"Use of satellite images for forest fires in area determination and monitoring,\" 2009 4th International Conference on Recent Advances in Space Technologies, Istanbul, Turkey, 2009, pp. 27-32.doi: 10.1109/RAST.2009.5158210 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5158210&isnumber=5158174 [3] C. Nithesh et al., \"RSA based Forest fire spread detection using Drones and Image Processing,\" 2022 IEEE 3rd Global Conference for Advancement in Technology (GCAT), Bangalore, India, 2022, pp. 1-4. doi: 10.1109/GCAT55367.2022.9972001 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9972001&isnumber=9971808 [4] C. Naimeng, Y. Wanjun and W. Xiaoyu, \"Smoke detection for early forest fire in aerial photography based on GMM background and wavelet energy,\" 2021 IEEE International Conference on Power Electronics, Computer Applications (ICPECA), Shenyang, China, 2021, pp. 763-765. doi: 10.1109/ICPECA51329.2021.9362647 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9362647&isnumber=9362511 [5] Li Jie and Xiao Jiang, \"Forest fire detection based on video multi-feature fusion,\" 2009 2nd IEEE International Conference on Computer Science and Information Technology, Beijing, China, 2009, pp. 19-22. doi: 10.1109/ICCSIT.2009.5234862 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5234862&isnumber=5234374 [6] W. Wang, Q. Huang, H. Liu, Y. Jia and Q. Chen, \"Forest Fire Detection Method Based on Deep Learning,\" 2022 International Conference on Cyber-Physical Social Intelligence (ICCSI), Nanjing, China, 2022, pp. 23-28. doi: 10.1109/ICCSI55536.2022.9970702 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9970702&isnumber=9970555 [7] T. P. D. Pieris, A. L. Kulasekera and K. V. D. S. Chathuranga, \"Multi-AI Based Wireless Sensor Node for Forest Fire Detection,\" 2022 7th International Conference on Information Technology Research (ICITR), Moratuwa, Sri Lanka, 2022, pp. 1-6. doi: 10.1109/ICITR57877.2022.9992560 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9992560&isnumber=9992370 [8] T. Wang, J. Hu, T. Ma and J. Song, \"Forest fire detection system based on Fuzzy Kalman filter,\" 2020 International Conference on Urban Engineering and Management Science (ICUEMS), Zhuhai, China, 2020, pp. 630-633. doi: 10.1109/ICUEMS50872.2020.00138 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9151570&isnumber=9151451 [9] W. S. Mseddi, R. Ghali, M. Jmal and R. Attia, \"Fire Detection and Segmentation using YOLOv5 and U-NET,\" 2021 29th European Signal Processing Conference (EUSIPCO), Dublin, Ireland, 2021, pp. 741-745. doi: 10.23919/EUSIPCO54536.2021.9616026 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9616026&isnumber=9615917, [10] J. Zhang, W. Li, Z. Yin, S. Liu and X. Guo, \"Forest fire detection system based on wireless sensor network,\" 2009 4th IEEE Conference on Industrial Electronics and Applications, Xi\'an, China, 2009, pp. 520-523. doi: 10.1109/ICIEA.2009.5138260 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5138260&isnumber=5138149

Copyright

Copyright © 2023 S. Sreekanth, Ch. Sriram, P. Sujan. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.