计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (4): 64-73.doi: 10.11896/jsjkx.241000093

• 智能嵌入式系统 • 上一篇    下一篇

渐进自适应特征融合的轻量化火焰检测算法研究

李啸澜, 马勇   

  1. 南京理工大学计算机科学与工程学院 南京 210094
  • 收稿日期:2024-10-18 修回日期:2025-02-14 出版日期:2025-04-15 发布日期:2025-04-14
  • 通讯作者: 马勇(mayong@njust.edu.cn)
  • 作者简介:(xiaolan.li.work@qq.com)
  • 基金资助:
    国家自然科学基金(61773210)

Study on Lightweight Flame Detection Algorithm with Progressive Adaptive Feature Fusion

LI Xiaolan, MA Yong   

  1. School of Computer Science and Engineering,Nanjing University of Science and Technology,Nanjing 210094,China
  • Received:2024-10-18 Revised:2025-02-14 Online:2025-04-15 Published:2025-04-14
  • About author:LI Xiaolan,born in 2001,postgraduate,is a student member of CCF (No.U9103G).His main research interests include target detection and edge computing.
    MA Yong,born in 1983,Ph.D,associate professor.His main research interests include intelligent robotics and intelligent hardware design.
  • Supported by:
    National Natural Science Foundation of China(61773210).

PDF (PC)

102

摘要： 针对视觉安防系统在边缘计算平台部署火焰检测模型时面临的精度与实时性难以平衡的问题,提出一种渐进自适应特征融合的轻量化火焰检测算法。首先,设计轻量级稀疏卷积算子降低模型计算复杂度与内存访问开销。其次,针对分组卷积的通道间信息交互缺陷,基于残差思想构建长距离上下文特征增强的轻量级特征提取组件。为解决深度骨干网络中特征丢失及背景干扰问题,创新性地提出基于高频增强的轻量级特征强化机制,优化空间域和通道域参数,缓解背景干扰问题。在此基础上,建立特征增强-渐进自适应特征融合框架,促进不同尺度特征图充分融合,提高特征图利用率,增强对多尺度目标的识别效果。实验结果表明,所提方法在实时推理速度最高达到27.1 FPS的同时,参数量降低至2.1×106,较基准模型减少69.5%,并达到83.4%的mAP@0.5检测精度,显著优于现有主流方法。

关键词: 深度学习, 计算机视觉, 目标检测, 轻量化神经网络, 特征提取网络, 特征融合网络, 特征增强

Abstract: In response to the challenge of balancing accuracy and real-time performance when deploying flame detection models on edge computing platforms for visual security systems,a lightweight flame detection algorithm featuring progressive adaptive feature fusion is proposed.Firstly,a lightweight sparse convolution operator is designed to reduce the model’s computational complexity and memory access cost.Subsequently,to address the shortcomings of inter-channel information exchange in grouped convolutions,a lightweight feature extraction component is constructed based on the residual concept,enhancing long-distance contextual features.To tackle the issues of feature loss and background interference in deep backbone networks,an innovative lightweight feature enhancement mechanism based on high-frequency augmentation is proposed,optimizing the parameters in both spatial and channel domains to mitigate background disturbances.On this basis,a feature enhancement-progressive adaptive feature fusion framework is established to facilitate the thorough integration of feature maps at different scales,thereby improving the utilization of feature maps and enhancing the recognition effectiveness of multi-scale targets.Experimental results demonstrate that this method achieves a real-time inference speed of up to 27.1 FPS,reduces the parameter count to 2.1 M,which is a 69.5% reduction compared to the baseline model,and attains a detection accuracy of 83.4% mAP@0.5,significantly outperforming existing mainstream methods.

Key words: Deep learning, Computer vision, Object detection, Lightweight neural network, Feature extraction network, Feature fusion network, Feature enhancement

中图分类号: 

  • TP391
[1]CHEN J,HE Y,WANG J.Multi-feature fusion based fast videoflame detection[J].Building and Environment,2010,45(5):1113-1122.
[2]DEVE K B,HANCKE G P,SILVA B J.Design of a smart fire detection system[C]//IECON 2016－42nd Annual Conference of the IEEE Industrial Electronics Society.IEEE,2016:6205-6210.
[3]WU Q Y,YAN Y Y,DU J,et al.Flame Detection AlgorithmBased on Multi-feature Fusion[J].Journal of Intelligent Systems,2015,10(2):240-247.
[4]CORTES C,VAPNIK V.Support-vector networks[J].Machine learning,1995,20:273-297.
[5]JIA Y,W H Q,HU Y,et al.Image-based Flame Recognition Based on Improved Hierarchical Clustering and SVM[J].Computer Engineering and Applications,2014 (5):165-168.
[6]EMMY PREMA C,VINSLEY S S,SURESH S.Efficient flame detection based on static and dynamic texture analysis in forest fire detection[J].Fire technology,2018,54:255-288.
[7]YAN Y Y,CHEN C X,LIU Y A,et al.Flame Detection of Dynamic Texture Features in Dimension-Weighted Mode[J].Chinese Journal of Intelligent Systems,2017,12(4):548-555.
[8]HOWARD A G,ZHU M,CHEN B,et al.Mobilenets:Efficient convolutional neural networks for mobile vision applications[J].arXiv:1704.04861,2017.
[9]SANDLER M,HOWARD A,ZHU M,et al.Mobilenetv2:In-verted residuals and linear bottlenecks[C]//2018 IEEE Confe-rence on Computer Vision and Pattern Recognition.IEEE,2018:4510-4520.
[10]HOWARD A,SANDLER M,CHU G,et al.Searching for mobilenetv3[C]//2019 IEEE/CVF International Conference on Computer Vision.IEEE,2019:1314-1324.
[11]ZHANG X,ZHOU X,LIN M,et al.Shufflenet:An extremelyefficient convolutional neural network for mobile devices[C]//2018 IEEE Conference on Computer Vision and Pattern.IEEE,2018:6848-6856.
[12]CHEN J,KAO S,HE H,et al.Run,Don't Walk:Chasing Higher FLOPS for Faster Neural Networks[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE,2023:12021-12031.
[13]HAN K,WANG Y,TIAN Q,et al.Ghostnet:More featuresfrom cheap operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE,2020:1580-1589.
[14]TAN M,CHEN B,PANG R,et al.Mnasnet:Platform-awareneural architecture search for mobile[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE,2019:2820-2828.
[15]CHEN Y,KALANTIDIS Y,LI J,et al.A^2-nets:Double attention networks[C]//Proceedings of the 32nd International Conference on Neural Information Processing Systems.2018:350-359.
[16]ZHAO H,JIA J,KOLTUN V.Exploring self-attention for image recognition[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:10076-10085.
[17]HU J,SHEN L,SUN G.Squeeze-and-excitation networks[C]//2018 IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2018:7132-7141.
[18]WANG Q,WU B,ZHU P,et al.ECA-Net:Efficient channel attention for deep convolutional neural networks[C]//Procee-dings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:11534-11542.
[19]YANG L,ZHANG R Y,LI L,et al.Simam:A simple,parameter-free attention module for convolutional neural networks[C]//2021 International Conference on Machine Learning.PMLR,2021:11863-11874.
[20]WOO S,PARK J,LEE J Y,et al.Cbam:Convolutional block attention module[C]//2018 European Conference on Computer Vision.Springer,2018:3-19.
[21]LIN T Y,DOLLAR P,GIRSHICK R,et al.Feature pyramidnetworks for object detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2017:2117-2125.
[22]LIU S,QI L,QIN H,et al.Path aggregation network for instance segmentation[C]//2018 IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2018:8759-8768.
[23]TAN M,PANG R,LE Q V.Efficientdet:Scalable and efficient object detection[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE,2020:10781-10790.
[24]GHIASI G,LIN T Y,LE Q V.Nas-fpn:Learning scalable feature pyramid architecture for object detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:7036-7045.
[25]LIU S,HUANG D,WANG Y.Learning spatial fusion for single-shot object detection[J].arXiv:1911.09516,2019.
[26]HARIHARAN B,MALIK J,RAMANAN D.Discriminativedecorrelation for clustering and classification[C]//European Conference on Computer Vision.Berlin:Springer,2012:459-472.
[27]HE K,ZHANG X,REN S,et al.Deep residual learning forimage recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:770-778.
[28]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[J].arXiv:1409.1556,2014.
[29]LI C,LI L,JIANG H,et al.YOLOv6:A single-stage object detection framework for industrial applications[J].arXiv:2209.02976,2022.
[30]MA N,ZHANG X,ZHENG H T,et al.Shufflenet v2:Practical guidelines for efficient cnn architecture design[C]//2018 European Conference on Computer Vision.Munich:Springer,2018:116-131.
[31]DAQUAN Z,HOU Q,CHEN Y,et al.Rethinking Bottleneck Structure for Efficient Mobile Network Design[J].arXiv:2007.02269,2020.
[32]HE Y,SAHMA A,HE X,et al.FireNet:A Lightweight and Ef-ficient Multi-Scenario Fire Object Detector[J].Remote Sensing,2024,16(21):4112.
[33]SHEN P,SUN N,HU K,et al.FireViT:An Adaptive Lightweight Backbone Network for Fire Detection[J].Forests,2023,14(11):2158.
[34]LIANG Y,CHEN T,ZHANG W.Multi-Scale Fire DetectionAlgorithm with Adaptive Attention[J].Transactions of Beijing institute of Technology,2024,44(1):91-101.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发