Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Laser & Optoelectronics Progress, Volume. 58, Issue 24, 2411003(2021)

Improved Flame Target Detection Algorithm Based on YOLOv3

Binbin Zhang and Zilai·mahemuti Pa*
Author Affiliations
  • School of Electrical Engineering, Xinjiang University, Urumqi, Xinjiang 830047, China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (10)
    Equations (0)
    References (19)
    Cited By (1)
    Tools
    Paper Information
    Topics
    Figures & Tables(10)
    Structure of the residual module

    Fig. 1. Structure of the residual module

    Download full size
    Structure of the YOLOv3 algorithm

    Fig. 2. Structure of the YOLOv3 algorithm

    Download full size
    Schematic diagram of anchor prediction of YOLOv3 algorithm

    Fig. 3. Schematic diagram of anchor prediction of YOLOv3 algorithm

    Download full size
    Structure of the large-scale GCM

    Fig. 4. Structure of the large-scale GCM

    Download full size
    Structure of the CSA module

    Fig. 5. Structure of the CSA module

    Download full size
    Network structure of our algorithm

    Fig. 6. Network structure of our algorithm

    Download full size
    Performance comparison between our algorithm and the YOLOv3 algorithm. (a) mAP; (b) F1 score; (c) Loss

    Fig. 7. Performance comparison between our algorithm and the YOLOv3 algorithm. (a) mAP; (b) F1 score; (c) Loss

    Download full size
    Detection results of different algorithms. (a) Original YOLOv3 algorithm; (b) our algorithm

    Fig. 8. Detection results of different algorithms. (a) Original YOLOv3 algorithm; (b) our algorithm

    Download full size

    • Table 1. Detection results of different algorithms

      View table

      Table 1. Detection results of different algorithms

      AlgorithmCategoryPrecision /%Recall /%F1 score /%mAP /%FPS
      YOLOv3fire91.7269.7679.0085.3526.9
      light82.1775.7179.00
      Retinanet[18]fire90.7178.4384.0086.2218.5
      light83.2181.4382.00
      Centernet-HG[19]fire93.5777.9585.0088.1915.3
      light88.4682.1485.00
      Oursfire95.0778.9086.0089.8220.2
      light89.3984.2987.00

    • Table 2. Ablation experiment results

      View table

      Table 2. Ablation experiment results

      NetworkFPNCSAGCMLossmAP /%
      1××××85.35
      2√×××87.80
      3√√××88.52
      4√√√×88.97
      5√√√√89.82
    Tools

    Get Citation

    Copy Citation Text

    Binbin Zhang, Zilai·mahemuti Pa. Improved Flame Target Detection Algorithm Based on YOLOv3[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2411003

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Set citation alerts for article
    Save article for my favorites
    Paper Information

    Category: Imaging Systems

    Received: Jan. 20, 2021

    Accepted: Mar. 4, 2021

    Published Online: Nov. 29, 2021

    The Author Email: Zilai·mahemuti Pa (294625876@qq.com)

    DOI:10.3788/LOP202158.2411003

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology