Research on fire detection algorithm based on improved YOLOv8
Today,the fire problem is a major disaster that people all over the world have to face.With the rapid development of the economy,the increasing social wealth,and the gradual expan-sion of the city scale,the importance of fire protection work has become more and more promi-nent.However,at present,the traditional relying on physical sensor equipment such as light,smoke or temperature sense for fire early warning detection,this method of information single leads to limited range,difficult to meet the real-time fire detection requirements in complex en-vironments,so the YOLOv8 network model is introduced to detect fire.This paper introduces the YOLOv8 algorithm and main structure,builds the experimental environment,annotates the pictures,establishes a self-made dataset,trains the algorithm on the dataset,predicts the trained model,analyzes the data through the experimental effect,and discusses the future development direction of fire protection technology in depth.
NETL
NSTL
万方数据
