Fire detection algorithm based on the fusion of YOLOv8 and Deformable Conv DCN

Lin Po Shang, Yan Zuo Chang, Yi Chen, Yong Shan Ou

Fire identification, Deep learning, YOLOv8, Deformable Conv

With the progress of fire monitoring and Coping technique, image recognition based on deep learning has shown great potential in the field of fire detection. Aiming at the accuracy and efficiency problems existing in the existing object detection algorithms, this study proposed an improved YOLOv8 algorithm to improve the real-time recognition capability in the fire scene. Through experimental verification on standard fire data sets, this study evaluated the detection performance of the improved YOLOV8 algorithm fused with Deformable Conv. The experimental results show that the improved YOLOv8 has improved the fire identification accuracy compared with the traditional version, and has certain potential for practical application in fire monitoring system.

Received: 29 Jun 2024, Receive in revised form: 31 Jul 2024, Accepted: 08 Aug 2024, Available online: 15 Aug 2024

10.22161/ijaers.118.2