Improved multi-scale flame detection method

Yi-cheng HOU; Hui-qin WANG; Ke WANG

doi:10.37188/CJLCD.2020-0221

您当前的位置：

首页 >

文章列表页 >

Improved multi-scale flame detection method

Image Processing | 更新时间：2021-04-28

- Improved multi-scale flame detection method
- Chinese Journal of Liquid Crystals and Displays Vol. 36, Issue 5, Pages: 751-759(2021)
- 作者机构：
  
  西安建筑科技大学信息与控制工程学院, 陕西西安 710055
- 作者简介：
- 基金信息：
  
  Supported by Shaanxi Provincial Department of Science and Technology International Science and Technology Cooperation Program Project(2020KW-012);Shaanxi Provincial Department of Education Key Project High-end Think Tank(18JT006);Xi'an Science and Technology Bureau Project(GXYD10.1)
- DOI：10.37188/CJLCD.2020-0221
  CLC： TP394.1;TH691.9
- Received：31 August 2020，
  
  Revised：11 November 2020，
  
  Published：2021-05
- 稿件说明：
移动端阅览
Yi-cheng HOU, Hui-qin WANG, Ke WANG. Improved multi-scale flame detection method[J]. Chinese journal of liquid crystals and displays, 2021, 36(5): 751-759.
DOI：

Yi-cheng HOU, Hui-qin WANG, Ke WANG. Improved multi-scale flame detection method[J]. Chinese journal of liquid crystals and displays, 2021, 36(5): 751-759. DOI： 10.37188/CJLCD.2020-0221.

摘要

网络层数的加深会造成对火焰目标深层特征细节信息表征能力减弱，同时提取了低相关度的冗余特征，导致火焰识别精度不高。针对该问题，提出了一种基于改进Faster R-CNN的火焰检测方法，以提高在深层网络下的火焰识别精度。首先利用ResNet50网络提取火焰特征，并添加SENet模块降低火焰目标冗余特征；然后将深层特征和浅层特征进行多尺度特征融合，增强深层特征的细节信息；最后训练网络，实现对火焰目标的识别定位。实验通过构建VOC火焰数据集进行网络训练，使用测试集进行检测，并进行特征图可视化对比，相比于改进前模型，本文模型平均精度提高了7.78%，召回率提高了9.05%，精确率提高了12.54%。本文提出的火焰目标检测模型，通过结合注意力机制模块和多尺度特征融合机制，能够有效进行火焰目标特征提取，火焰目标的检测结果更加准确。

Abstract

The deepening of the number of network layers can weaken the ability to characterize the detailed information of the deep features of the flame target

and at the same time extract redundant features with low correlation

resulting in low flame recognition accuracy. Aiming at this problem

a flame detection method based on improved Faster R-CNN is proposed to improve the accuracy of flame recognition in deep networks. Firstly

the ResNet50 network is used to extract flame features

and the SENet module is added to reduce the redundant features of flame targets. Then

the deep features and shallow features are multi-scale feature fusion to enhance the detailed information of deep features. Finally

the network is trained to realize the recognition of flame targets positioning. In the experiment

the VOC flame data set is constructed for network training

the test set is used for detection

and the feature map visualization is compared. Compared with the model before the improvement

the AP value increases by 7.78%

the recall increases by 9.05%

and the precision increases by 12.54%. By combining the attention mechanism module and the multi-scale feature fusion mechanism

the flame target detection model proposed in this paper

can effectively extract the flame target feature

and the flame target detection result is more accurate.

关键词

Keywords

references

黄博 , 江慎旺 , 张增 , 等 . 自适应特征引流管故障智能识别方法 . 中国光学 , 2017 . 10 ( 3 ): 340 - 347 . https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201703006.htm https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201703006.htm .

B HUANG , S W JIANG , Z ZHANG , 等 . Intelligent identification algorithm of adaptive feature drainage tube fault . China Optics , 2017 . 10 ( 3 ): 340 - 347 . https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201703006.htm https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201703006.htm .

耿庆田 , 赵浩宇 , 于繁华 , 等 . 基于改进HOG特征提取的车型识别算法 . 中国光学 , 2018 . 11 ( 2 ): 174 - 181 . https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201802003.htm https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201802003.htm .

Q T GENG , H Y ZHAO , F H YU , 等 . Vehicle type recognition algorithm based on improved HOG feature . China Optics , 2018 . 11 ( 2 ): 174 - 181 . https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201802003.htm https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201802003.htm .

张健 , 钟中志 , 柯艳国 , 等 . 基于超像素和DTCWT的火焰检测 . 信息技术 , 2019 . 43 ( 6 ): 31 - 34 . https://www.cnki.com.cn/Article/CJFDTOTAL-HDZJ201906008.htm https://www.cnki.com.cn/Article/CJFDTOTAL-HDZJ201906008.htm .

J ZHANG , Z Z ZHONG , Y G KE , 等 . Flame detection based on super-pixel and DTCWT . Information Technology , 2019 . 43 ( 6 ): 31 - 34 . https://www.cnki.com.cn/Article/CJFDTOTAL-HDZJ201906008.htm https://www.cnki.com.cn/Article/CJFDTOTAL-HDZJ201906008.htm .

刘凯 , 魏艳秀 , 许京港 , 等 . 基于计算机视觉的森林火灾识别算法设计 . 森林工程 , 2018 . 34 ( 4 ): 89 - 95 . DOI: 10.3969/j.issn.1006-8023.2018.04.015 http://doi.org/10.3969/j.issn.1006-8023.2018.04.015 .

K LIU , Y X WEI , J G XU , 等 . Design of forest fire identification algorithm based on computer vision . Forest Engineering , 2018 . 34 ( 4 ): 89 - 95 . DOI: 10.3969/j.issn.1006-8023.2018.04.015 http://doi.org/10.3969/j.issn.1006-8023.2018.04.015 .

张杰 , 隋阳 , 李强 , 等 . 基于卷积神经网络的火灾视频图像检测 . 电子技术应用 , 2019 . 45 ( 4 ): 34 - 38, 44 . https://www.cnki.com.cn/Article/CJFDTOTAL-DZJY201904008.htm https://www.cnki.com.cn/Article/CJFDTOTAL-DZJY201904008.htm .

J ZHANG , Y SUI , Q LI , 等 . Fire video image detection based on convolutional neural network . Application of Electronic Technique , 2019 . 45 ( 4 ): 34 - 38, 44 . https://www.cnki.com.cn/Article/CJFDTOTAL-DZJY201904008.htm https://www.cnki.com.cn/Article/CJFDTOTAL-DZJY201904008.htm .

蔡敏. 基于视频分析的森林烟火识别算法研究[D]. 南京: 东南大学, 2018.

CAI M. Research on forest smoke detection based on video analysis[D]. Nanjing: Southeast University, 2018. (in Chinese)

刘宇欣. 基于人工蜂群算法的矿用带式输送机早期火灾检测[D]. 西安: 西安科技大学, 2018.

LIU Y X. Research on early fire monitoring of mine belt conveyor based on artificial bee colony algorithm[D]. Xi'an: Xi'an University of Science and Technology, 2018. (in Chinese)

苗续芝. 基于视频图像的火灾检测研究与实现[D]. 徐州: 中国矿业大学, 2018.

MIAO X Z. Study and realization of fire detection based on video image[D]. Xuzhou: China University of Mining and Technology, 2018. (in Chinese)

王中林. 基于双目视觉的火焰目标检测[D]. 北京: 北京交通大学, 2018.

WANG Z L. Flame target detection based on binocular stereo vision[D]. Beijing: Beijing Jiaotong University, 2018. (in Chinese)

唐悦 , 吴戈 , 朴燕 . 改进的GDT-YOLOV 3目标检测算法 . 液晶与显示 , 2020 . 35 ( 8 ): 852 - 860 . http://cjlcd.lightpublishing.cn/thesisDetails#10.37188/YJYXS20203508.0852 http://cjlcd.lightpublishing.cn/thesisDetails#10.37188/YJYXS20203508.0852 .

Y TANG , G WU , Y PIAO . Improved algorithm of GDT-YOLOV 3 image target detection . Chinese Journal of Liquid Crystals and Displays , 2020 . 35 ( 8 ): 852 - 860 . http://cjlcd.lightpublishing.cn/thesisDetails#10.37188/YJYXS20203508.0852 http://cjlcd.lightpublishing.cn/thesisDetails#10.37188/YJYXS20203508.0852 .

刘嘉政 , 王雪峰 , 王甜 . 基于深度学习的树种图像自动识别 . 南京林业大学学报(自然科学版) , 2020 . 44 ( 1 ): 138 - 144 . https://www.cnki.com.cn/Article/CJFDTOTAL-NJLY202001020.htm https://www.cnki.com.cn/Article/CJFDTOTAL-NJLY202001020.htm .

J Z LIU , X F WANG , T WANG . Automatic identification of tree species based on deep learning . Journal of Nanjing Forestry University (Natural Science Edition) , 2020 . 44 ( 1 ): 138 - 144 . https://www.cnki.com.cn/Article/CJFDTOTAL-NJLY202001020.htm https://www.cnki.com.cn/Article/CJFDTOTAL-NJLY202001020.htm .

丁立德 , 胡怀湘 . 基于FPGA的CNN应用加速技术 . 信息技术 , 2019 . 43 ( 12 ): 110 - 115 . DOI: 10.3969/j.issn.1674-2117.2019.12.037 http://doi.org/10.3969/j.issn.1674-2117.2019.12.037 .

L D DING , H X HU . An acceleration technique for CNN application based on FPGA . Information Technology , 2019 . 43 ( 12 ): 110 - 115 . DOI: 10.3969/j.issn.1674-2117.2019.12.037 http://doi.org/10.3969/j.issn.1674-2117.2019.12.037 .

任嘉锋 , 熊卫华 , 吴之昊 , 等 . 基于改进YOLOv3的火灾检测与识别 . 计算机系统应用 , 2019 . 28 ( 12 ): 171 - 176 . https://www.cnki.com.cn/Article/CJFDTOTAL-XTYY201912026.htm https://www.cnki.com.cn/Article/CJFDTOTAL-XTYY201912026.htm .

J F REN , W H XIONG , Z H WU , 等 . Fire detection and identification based on improved YOLOv3 . Computer Systems & Applications , 2019 . 28 ( 12 ): 171 - 176 . https://www.cnki.com.cn/Article/CJFDTOTAL-XTYY201912026.htm https://www.cnki.com.cn/Article/CJFDTOTAL-XTYY201912026.htm .

徐登 , 黄晓东 . 基于改进双流卷积网络的火灾图像特征提取方法 . 计算机科学 , 2019 . 46 ( 11 ): 291 - 296 . DOI: 10.11896/jsjkx.180901640 http://doi.org/10.11896/jsjkx.180901640 .

D XU , X D HUANG . Fire images features extraction based on improved two-stream convolution network . Computer Science , 2019 . 46 ( 11 ): 291 - 296 . DOI: 10.11896/jsjkx.180901640 http://doi.org/10.11896/jsjkx.180901640 .

段锁林 , 刘福 , 高仁洲 , 等 . 基于卷积神经网络的火焰识别 . 计算机工程与设计 , 2019 . 40 ( 11 ): 3288 - 3292, 3298 . https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201911044.htm https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201911044.htm .

S L DUAN , F LIU , R Z GAO , 等 . Flame recognition based on convolution neural network . Computer Engineering and Design , 2019 . 40 ( 11 ): 3288 - 3292, 3298 . https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201911044.htm https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201911044.htm .

回天 , 哈力旦·阿布都热依木 , 杜晗 . 结合Faster R-CNN的多类型火焰检测 . 中国图象图形学报 , 2019 . 24 ( 1 ): 73 - 83 . https://www.cnki.com.cn/Article/CJFDTOTAL-ZGTB201901008.htm https://www.cnki.com.cn/Article/CJFDTOTAL-ZGTB201901008.htm .

T HUI , HALIDAN·ABUDUREYIMU , H DU . Multi-type flame detection combined with Faster R-CNN . Journal of Image and Graphics , 2019 . 24 ( 1 ): 73 - 83 . https://www.cnki.com.cn/Article/CJFDTOTAL-ZGTB201901008.htm https://www.cnki.com.cn/Article/CJFDTOTAL-ZGTB201901008.htm .

S Q REN , K M HE , R GIRSHICK , 等 . Faster R-CNN: towards real-time object detection with region proposal networks . IEEE Transactions on Pattern Analysis and Machine Intelligence , 2017 . 39 ( 6 ): 1137 - 1149 . http://europepmc.org/abstract/MED/27295650 http://europepmc.org/abstract/MED/27295650 .

HE K M, ZHANG X Y, REN S Q, et al . Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition ( CVPR ). Las Vegas, NV, USA: IEEE, 2016: 770-778.

J HU , L SHEN , S ALBANIE , 等 . Squeeze-and-excitation networks . IEEE Transactions on Pattern Analysis and Machine Intelligence , 2020 . 42 ( 8 ): 2011 - 2023 .

A KRIZHEVSKY , I SUTSKEVER , G E HINTON . ImageNet classification with deep convolutional neural networks . Communications of the ACM , 2017 . 60 ( 6 ): 84 - 90 . http://users.ics.aalto.fi/perellm1/thesis/summaries_html/node64.html http://users.ics.aalto.fi/perellm1/thesis/summaries_html/node64.html .

T Y CHUANG , J Y HAN , D J JHAN , 等 . Geometric recognition of moving objects in monocular rotating imagery using faster R-CNN . Remote Sensing , 2020 . 12 ( 12 ): 1908 http://www.researchgate.net/publication/342175133_Geometric_Recognition_of_Moving_Objects_in_Monocular_Rotating_Imagery_Using_Faster_R-CNN http://www.researchgate.net/publication/342175133_Geometric_Recognition_of_Moving_Objects_in_Monocular_Rotating_Imagery_Using_Faster_R-CNN .

HUANG G, LIU Z, VAN DER MAATEN L, et al . Densely connected convolutional networks[J]. arXiv : 1608.06993, 2016.

REDMON J, FARHADI A. YOLOv3: an incremental improvement[J]. arXiv : 1804.02767, 2018.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Multi-scale pedestrian detection algorithm based on joint head and overall information

Detection and segmentation method of surgical instruments based on improved YOLOv5s

Infrared dim-small target detection under complex background based on attention mechanism

Semantic segmentation algorithm based on class feature attention mechanism fusion

Related Author

Yi-cheng HOU

Hui-qin Wang

Ke Wang

MA Ximing

LI Ning

WU Di

WANG Jianfei

YU Xiangyue

Related Institution

College of information and control engineering， Xi'an University of Architecture and Technology.， Xian

Military Representative Office Stationed in a Region by Air Force Equipment Department

Changchun Institute of Optics， Fine Mechanics and Physics， Chinese Academy of Sciences

School of Computer Science and Technology， Shandong University of Technology

University of Chinese Academy of Sciences

AI问答

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China 130033 Postal code：130033
Tel：0431-86176059 Email：yjxs@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰