主管:中华人民共和国应急管理部
主办:应急管理部天津消防研究所
ISSN 1009-0029  CN 12-1311/TU

消防科学与技术 ›› 2023, Vol. 42 ›› Issue (1): 42-46.

• • 上一篇    下一篇

基于轻量化CNN的船舶火灾实时探测算法研究

刘义琛1 ,张 彬1 ,王学岐2 ,童家鹏1
  

  1. (1. 大连海事大学 轮机工程学院,辽宁 大连 116026;2. 中国石油集团安全环保技术研究院有限公司,辽宁 大连 116031)
  • 出版日期:2023-01-15 发布日期:2023-01-15
  • 作者简介:刘义琛(1999- ),男,河南焦作人,大连海事大学轮机工程学院硕士研究生,主要从事基于机器视觉的船舶火灾安全研究,辽宁省大连市凌水路1号大连海事大学轮机综合试验室,116026。
  • 基金资助:
    辽宁省自然科学基金项目(2020JH/10300107);国家自然科学基金项目(51306026);中央高校基本科研业务费专项资金项目(3132019038,3132019339)

Research on real-time detection algorithm of ship fire based on lightweight CNN

Liu Yichen1 ,Zhang Bin1 ,Wang Xueqi2 ,Tong Jiapeng1
  

  1. (1. Marine Engineering College, Dalian Maritime University, Liaoning Dalian 116026, China; 2. CNPC Research Institute of Safety & Environment Technology, Liaoning Dalian 116031, China)
  • Online:2023-01-15 Published:2023-01-15

摘要:

针对船舶火灾探测快速、准确及实时的实际工程需求,提出一种基于改进 YOLOv5s 的轻量化、高精度的船舶火灾探测 SG-YOLO 算法。使用融合无参数注意力机制的 GhostNet 卷积结构实现算法轻量化,引入二维注意力机制与加权双向特征金字塔结构增强火焰特征提取能力,有效解决远景密集小尺寸火焰以及目标框定位不精准的问题。在自建火灾集对比试验中,与 YOLOv5s 6.0 相比,该算法模型参数量减少 46.2%,探测速度提升 38.7%,达到 86 f/s,探测精度提升 0.9%。

关键词:

Abstract:

For the actual engineering requirements of fast, accurate and real-time detection of ship fire, a lightweight and highprecision SG-YOLO algorithm based on improved YOLOv5s is proposed. The GhostNet convolution structure fused with the parameter- free attention mechanism is used to achieve algorithm lightweight, the 2D attention mechanism and the bi- directional feature pyramid network are introduced to enhance the feature extraction ability of flame, and solve the problems of dense and small flames and imprecise position of target frame. In the self built fire set comparison experiment, compared with YOLOv5s 6.0, the model parameters is reduced by 46.2% , the detection speed is improved by 38.7%, reaching 86 f/s, and the detection accuracy is improved by 0.9%.

Key words: