IMPROVEMENT OF THE YOLOV8 MODEL FOR REMOTE SENSING OF RAPID DESTRUCTIVE PROCESSES
DOI:
https://doi.org/10.35546/kntu2078-4481.2025.1.2.18Keywords:
remote sensing, unmanned aerial vehicles, rapid destructive processes, forest fire, combustion area, attention mechanism, feature identification, object recognition, imagesAbstract
Timely and accurate detection of forest fires as the most common class of rapid destructive processes is crucial for stopping them and minimizing their consequences. Remote sensing technologies from unmanned aerial vehicles, machine learning, and computer vision can be used for this purpose. However, the influence of the external environment and several uncertainties, distortions, and motion dynamics create problems in identifying fire signs, as well as computational complexity hinder the operation of recognition algorithms in real time. To solve these problems, the paper proposes a “lightweight” model for recognizing forest fire sources during remote sensing, YOLOv8N, which improves the basic YOLOv8n model by using the GhostNetv2 module at the backbone level together with the DFC attention module instead of the traditional convolution operation, which allows significantly reducing the number of model parameters while maintaining its performance, and the MHSA attention mechanism in C2f operations, which improves the ability to obtain signs of burning sources and increases the accuracy of detecting small burning areas. Also, at the intermediate level of the model, the SegNeXt self-attention mechanism is used in C2f operations, which allows to increase the accuracy of detecting fire signs in difficult conditions. The YOLOv8N model increases the accuracy, completeness, harmonic mean, and precision by 4.3 %, 7.5 %, 4.8 % and 5.9 % respectively compared to the base model YOLOv8n, the number of parameters is also reduced by 33.3 %. Therefore, the proposed model provides a high level of accuracy in detecting forest fire features, while maintaining a balance between computational complexity and model efficiency, which guarantees its ability to work in remote sensing systems from unmanned aerial vehicles in real time.
References
Hillayová, M., Holécy, J., Korísteková, K., Bakšová, M., Ostrihoň, M., Škvarenina, J.: Ongoing Climatic Change Increases the Risk of Wildfires. Case Study: Carpathian Spruce Forests. Journal of Environmental Management, 2023, vol. 337, pp. 117620.
Stocks, B., Mason, J., Todd, J., Bosch, E. M., Wotton, B. M., Amiro, B. D., Flannigan, M. D., Hirsch, K. G., Logan, K. A., Martell, D. L.: Large Forest Fires in Canada, 1959–1997. J. of Geographical Research, 2002, vol. 107, 8149.
Howell, A. N., Belmont, E. L., McAllister, S. S., Finney, M. A.: An Investigation of Oxygen Availability in Spreading Fires. Fire Technology, 2023, vol. 59, pp. 2147–2176.
Sherstjuk, V., Zharikova, M., Dorovskaja, I., Sheketa, V.: Assessing Forest Fire Dynamics in UAV-Based Tactical Monitoring System. Advances in Intelligent Systems and Computing, 2021, vol. 1246, pp. 285–301.
Sherstjuk, V., Zharikova, M., Dorovskaja, I.: 3D Fire Front Reconstruction in UAV-Based Forest-Fire Monitoring System. Data Stream Mining & Processing: Proc. of IEEE Third Int. Conf. DSMP’2020, Lviv, 2020, pp. 243–248.
Bessho, K., Date, K., Hayashi, M., Ikeda, A., Imai, T., Inoue, H., Kumagai, Y., Miyakawa, T., Murata, H., Ohno, T.: An Introduction to Himawari-8/9 Japan’s New-Generation Geostationary Meteorological Satellites. Journal of the Meteorological Society of Japan, Ser. II 2016, vol. 94, pp. 151–183.
Justice, C. O., Giglio, L., Korontzi, S., Owens, J., Morisette, J. T., Roy, D., Descloitres, J., Alleaume, S., Petitcolin, F., Kaufman, Y.: The MODIS Fire Products. Remote sensing of Environment, 2002, vol. 83(1–2), pp. 244–262.
Sherstjuk, V., Zharikova, M., Dorovskaja, I., Chornyi, D., Gusev, V., Sokol, I.: 3D Fire Front Reconstruction Based on Multi-View Observation and Complex Uncertainty Model. Advanced Computer Information Technologies: Proc. of 11th Int. Conf., Deggendorf, Germany, 2021, pp. 761–765.
Osco, L. P., Junior, J. M., Ramos, A. P. M., de Castro Jorge, L. A., Fatholahi, S. N., de Andrade Silva, J., Matsubara, E. T., Pistori, H., Gonçalves, W. N., Li, J.: A Review on Deep Learning in UAV Remote Sensing. International Journal of Applied Earth Observation and Geoinformation, 2021, vol. 102, 102456.
Li, C., Li, G., Song, Y., He, Q., Tian, Z., Xu, H., Liu, X.: Fast Forest Fire Detection and Segmentation Application for UAV-Assisted Mobile Edge Computing System. IEEE Internet of Things Journal, 2023, vol.11(16), pp. 26690–26699.
Yang, X., Hua, Z., Zhang, L., Fan, X., Zhang, F., Ye, Q., Fu, L.: Preferred Vector Machine for Forest Fire Detection. Pattern Recognition, 2023, vol. 143, 109722.
Maeda, N., Tonooka, H.: Early Stage Forest Fire Detection from Himawari-8 AHI Images Using a Modified MOD14 Algorithm Combined with Machine Learning. Sensors, 2022, vol. 23, 210.
Sathishkumar, V. E., Cho, J., Subramanian, M., Naren, O. S.: Forest Fire and Smoke Detection Using Deep Learning-Based Learning without Forgetting. Fire Ecology, 2023, vol. 19, p. 9.
Liu, F., Chen, R., Zhang, J., Xing, K., Liu, H., Qin, J.: R2YOLOX: A Lightweight Refined Anchor-Free Rotated Detector for Object Detection in Aerial Images. IEEE Transactions on Geoscience and Remote Sensing, 2022, vol. 60, pp. 1–15.
Guan, R., Li, Z., Tu, W., Wang, J., Liu, Y., Li, X., Tang, C., Feng, R.: Contrastive Multi-View Subspace Clustering of Hyperspectral Images Based on Graph Convolutional Networks. IEEE Transactions on Geoscience and Remote Sensing, 2024, vol. 62, pp. 1–14.
Lv, Q., Quan, Y., Sha, M., Feng, W., Xing, M.: Deep Neural Network-Based Interrupted Sampling Deceptive Jamming Countermeasure Method. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, vol. 15, pp. 9073–9085.
Li, Y., Zhang, S., Wang, W.-Q.: A Lightweight Faster R-CNN for Ship Detection in SAR Images. IEEE Geoscience and Remote Sensing Letters, 2020, vol. 19, pp. 1–5.
Zhang, W., Liu, Z., Zhou, S., Qi, W., Wu, X., Zhang, T., Han, L.: LS-YOLO: A Novel Model for Detecting Multi- Scale Landslides with Remote Sensing Images. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 1–14.
Zhang, L., Wang, M., Ding, Y., Bu, X.: MS-FRCNN: A Multi-Scale Faster RCNN Model for Small Target Forest Fire Detection. Forests, 2023, vol. 14(3), 616.
Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: You Only Look Once: Unified, Real-Time Object Detection. Arxiv, 2016, 1506.02640.
Niu, Z., Zhong, G., Yu, H.: A Review on the Attention Mechanism of Deep Learning. Neurocomputing, 2021, vol. 452, pp. 48–62.
Luo, M., Xu, L., Yang, Y., Cao, M., Yang, J.: Laboratory Flame Smoke Detection Based on an Improved YOLOX Algorithm. Applied Sciences, 2022, vol. 12(24), 12876.
Xue, Q., Lin, H., Wang, F.: FCDM: An Improved Forest Fire Classification and Detection Model Based on YOLOv5. Forests, 2022, vol. 13(12), 2129.
Chen, G., Cheng, R., Lin, X., Jiao, W., Bai, D., Lin, H.: LMDFS: A Lightweight Model for Detecting Forest Fire Smoke in UAV Images Based on YOLOv7. Remote Sensing, 2023, vol. 15(15), 3790.
Wang, Z., Hua, Z., Wen, Y., Zhang, S., Xu, X., Song, H.: E-YOLO: Recognition of Estrus Cow Based on Improved YOLOv8n Model. Expert Systems with Applications, 2024, vol. 238, 122212.
Mahasin, M., Dewi, I. A.: Comparison of CSPDarkNet53, CSPResNeXt-50, and EfficientNet-B0 Backbones on YOLO V4 as Object Detector. International Journal of Engineering, Science and Information Technology, 2022, vol 2(3), pp. 64–72.
Yi, H., Liu, B., Zhao, B., Liu, E.: Small Object Detection Algorithm Based on Improved YOLOv8 for Remote Sensing. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, vol. 17, pp. 1734–1747.
Neubeck, A., Van Gool, L.: Efficient Non-Maximum Suppression. Pattern recognition: Proc. of the 18th Int. Conf. ICPR’06, Hong Kong, China, 2006, vol. 3, pp. 850–855.
Tang, Y., Han, K., Guo, J., Xu, C., Xu, C., Wang, Y.: GhostNetv2: Enhance Cheap Operation with Long-Range Attention. Advances in Neural Information Processing Systems, 2022, vol. 35, pp. 9969–9982.
Zhang, Z.: Drone-YOLO: An Efficient Neural Network Method for Target Detection in Drone Images. Drones, 2023, 7(8), 526.
Zhao, F., Feng, F., Ye, S., Mao, Y., Chen, X., Li, Y., Ning, M., Zhang, M.: Multi-head self-attention mechanism- based global feature learning model for ASD diagnosis. Biomedical Signal Processing and Control, 2024, vol. 91, 106090.
Zhao, Z., Du, J., Li, C., Fang, X., Xiao, Y., Tang, J.: Dense Tiny Object Detection: A Scene Context Guided Approach and a Unified Benchmark. IEEE Transactions on Geoscience and Remote Sensing, 2024, vol. 62, pp. 1–13.
Wang, Z., Meng, Q., Tang, F., Qi, Y., Li, B., Liu, X., Kong, S., Li, X.: MSG-YOLO: A Lightweight Detection Algorithm for Clubbing Finger Detection. Electronics, 2024, vol. 13(22), 4549.
Guo, M.-H., Lu, C.-Z., Hou, Q., Liu, Z.-N., Cheng, M.-M., Hu, S.-M.: SegNeXt: rethinking convolutional attention design for semantic segmentation. Neural Information Processing Systems: Proc. of the 36th Int. Conf. NIPS’22, Red Hook, NY, USA, 2022, vol. 84, pp. 1140–1156.
Guo, M., Lu, C., Hou, Q., Liu, Z., Cheng, M., Hu, S.: SegNeXt: Rethinking Convolutional Attention Design for Semantic Segmentation. ArXiv, 2022, abs/2209.08575.
Sherstjuk, V., Zharikova, M.: Fire-Front Recognition in UAV-Based Forest-Fire Monitoring System Using Fuzzy Rough Soft Sets. Electrical and Computer Engineering: Proc. of IEEE 2nd Ukraine Conference UKRCON, Lviv, Ukraine, 2019, pp. 1091–1096.
Sherstjuk, V., Zharikova, M.: Evaluation of Fire Intensity Based on Neural Networks in a Forest-Fire Monitoring System. Electronics and Nanotechnology: Proc. of IEEE 39th Int. Conf. ELNANO’2019, Kyiv, Ukraine, 2019, pp. 802–807.
Wang, G., Li, H., Li, P., Lang, X., Feng, Y., Ding, Z., Xie, S.: M4SFWD: A Multi-Faceted Synthetic Dataset for Remote Sensing Forest Wildfires Detection. Expert Systems with Applications, 2024, vol. 245, 123489.
