基于自适应高斯混合模型与ResDN的火焰检测算法
DOI:
作者:
作者单位:

大连海事大学船舶电气工程学院

作者简介:

通讯作者:

中图分类号:

TP391.41

基金项目:

国家自然科学基金项目(面上项目,重点项目,重大项目)


Fire Detection Algorithm Combining Adaptive Gaussian Mixture Model and Residual Dense Network
Author:
Affiliation:

1.School of Marine Electrical Engineering,Dalian Maritime University,Dalian Liaoning 116026;2.China

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对火焰检测算法在复杂场景下误检率高、算法适应性差、效率低等问题,设计了一种轻量高效的两阶段视频火焰检测算法。第一阶段采用改进的自适应高斯混合模型(Adaptive Gaussian Mixture Model,AGMM)对视频图像序列进行快速背景建模,利用火焰的闪烁和涌动特性,提取出序列中的可疑候选区域。第二阶段使用残差深度归一化卷积神经网络 (Residual Deep Normalization And Convolutional Neural Network, ResDN)对可疑候选区域进行判别,并引入简化的残差块替换原有的卷积层进行轻量化设计,实现对火焰的检测与定位。相比于传统分类算法,所设计的两阶段视频火焰检测算法能够有效克服复杂场景下的环境干扰,准确快速地识别火焰,具有更高的检测率和适应性。

    Abstract:

    : A lightweight and efficient two-stage video flame detection algorithm is designed to address the high false positive rate, poor adaptability, and low efficiency of flame detection algorithms in complex scenes. In the first stage, an improved Adaptive Gaussian Mixture Model (AGMM) is used to rapidly construct the background model for the video image sequence. By exploiting the flickering and surging characteristics of flames, suspicious candidate regions are extracted from the sequence. In the second stage, a Residual Deep Normalization and Convolutional Neural Network (ResDN) is employed to discriminate the suspicious candidate regions. Additionally, a simplified residual block is introduced to replace the original convolutional layer, enabling lightweight design and achieving accurate flame detection and localization. Compared to traditional classification algorithms, the proposed two-stage video flame detection algorithm can effectively overcome environmental interference in complex scenes, accurately and rapidly identify flames, and exhibit higher detection rates and adaptability.

    参考文献
    相似文献
    引证文献
引用本文

王文标,时启衡,郝友维. 基于自适应高斯混合模型与ResDN的火焰检测算法[J]. 科学技术与工程, , ():

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2023-12-08
  • 最后修改日期:2024-06-06
  • 录用日期:2024-07-09
  • 在线发布日期:
  • 出版日期:
×
亟待确认版面费归属稿件,敬请作者关注
《科学技术与工程》入选维普《中文科技期刊数据库》自然科学类期刊月度下载排行榜TOP10