基于YOLOv8火灾检测及告警系统的设计与实现  

Design and Implementation of Fire Detection and Alarm System Based on YOLOv8

作　　者：朱贺新[1] 赵元苏 郭蕊[1] 赵志源 ZHU Hexin;ZHAO Yuansu;GUO Rui;Zhao Zhiyuan(School of Information Engineering,Beijing Polytechnic College,Beijing 100042,China;China Telecom Corporation Limited Chongqing Branch,Chongqing 401121,China)

机构地区：[1]北京工业职业技术学院信息工程学院,北京100042 [2]中国电信股份有限公司重庆分公司,重庆401121

出　　处：《北京工业职业技术学院学报》2025年第2期16-20,共5页Journal of Beijing Polytechnic College

基　　金：2024年北京工业职业技术学院科研课题(BGY2024KY-23)。

摘　　要：经过近十年迭代优化,YOLO模型已成为实时目标检测领域典型代表之一。将YOLOv8应用于火焰烟雾训练,能够在复杂环境中有效检测火焰和烟雾,为火灾预防和早期干预提供技术支持。基于YOLOv8设计的火灾检测与告警系统,包括数据采集、预处理、AI检测处理、告警联动、用户界面和审计等模块,能够完成火灾检测、邮件告警、短信通知等任务,具有一定的实用价值和推广前景。After nearly a decade of iterative optimization,the YOLO model has become one of the typical representatives in the field of real-time object detection.Applying YOLOv8 to flame and smoke training can effectively detect flames and smoke in complex environments,providing technical support for fire prevention and early intervention.The fire detection and alarm system designed based on YOLOv8 includes modules such as data collection,preprocessing,AI detection and processing,alarm linkage,user interface,and audit.It can complete tasks such as fire detection,email alerts,and SMS notifications,and has certain practical value and promotion prospects.

关 键 词：火灾检测 嵌入式 模型训练 告警联动 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]