Laser & Optoelectronics Progress, Volume. 59, Issue 12, 1210010(2022)

Registration Method for Power Equipment Infrared and Visible Images Based on Improved Curvature Scale Space Algorithm

Yunhong Li*, Xuemin Luo, Xueping Su, Yaolin Zhu, Lan Yao, and Jiaojiao Duan
Author Affiliations
  • School of Electronics and Information, Xi’an Polytechnic University, Xi’an 710048, Shaanxi , China
  • show less
    Figures & Tables(7)
    Freeman 8-adjacent direction value and chain code direction. (a) Freeman 8 direction value; (b) Freeman 8 chain code points
    Flow chart of improved CSS feature point detection algorithm
    Schematic diagram of distance from point to string
    Schematic diagram of feature main direction
    Infrared and visible image registration of power equipment. (a) Infrared images; (b) visible images; (c) registration images; (d) fusion images
    • Table 1. Comparison of registration results of different registration algorithms

      View table

      Table 1. Comparison of registration results of different registration algorithms

      SampleAlgorithmMatching logarithmCorrect logarithmPrecisionRecallRMSE
      Porcelain insulatorSURF241980.1%16.57%1.1637
      SIFT272178.8%18.49%1.2254
      CSS241979.2%22.89%0.6124
      Proposed221986.4%25.33%0.5382
      Surge arresterSURF706289.2%29.03%1.3783
      SIFT747297.2%33.51%1.0759
      CSS7272100.0%37.63%0.3728
      Proposed717098.6%35.45%0.3732
      Current transformerSURF171482.4%17.28%1.1639
      SIFT181266.2%5.87%2.2933
      CSS151066.7%6.49%1.6206
      Proposed231778.3%14.40%0.8547
      Oil-filled transformerSURF1048783.7%10.82%8.9042
      SIFT11210089.6%10.13%7.9943
      CSS998989.9%11.03%3.5009
      Proposed847791.7%11.32%3.3786
      Isolating switchSURF12510281.3%12.58%8.6207
      SIFT13112293.1%11.75%9.8135
      CSS12912395.3%18.36%1.7532
      Proposed12411995.6%19.10%1.6254
      Overhead networkSURF13710878.8%7.73%7.9346
      SIFT14411982.6%7.45%7.8762
      CSS1039491.3%12.81%1.7383
      Proposed1009393.0%13.13%1.6997
      Poles and towersSURF1135145.1%3.67%17.0624
      SIFT1274233.1%2.92%21.3027
      CSS907684.4%6.26%3.1172
      Proposed413687.8%10.17%2.9451
      OthersSURF381744.7%2.49%13.5783
      SIFT431432.6%1.94%16.0831
      CSS262180.8%3.85%1.7036
      Proposed222090.9%5.18%1.4723
    • Table 2. Comparison of registration time of different registration algorithms

      View table

      Table 2. Comparison of registration time of different registration algorithms

      AlgorithmSample
      Porcelain insulatorSurge arresterCurrent transformerOil-filled transformerIsolating switchOverhead networkPoles and towersOthers
      SURF1.39241.73231.23297.57417.21397.894314.230611.9724
      SIFT1.36621.72581.60038.08347.89418.220517.525615.0933
      CSS1.39051.70491.54526.79796.08916.769710.87876.0108
      Proposed1.29421.70271.18026.62235.77836.12679.21644.9132
    Tools

    Get Citation

    Copy Citation Text

    Yunhong Li, Xuemin Luo, Xueping Su, Yaolin Zhu, Lan Yao, Jiaojiao Duan. Registration Method for Power Equipment Infrared and Visible Images Based on Improved Curvature Scale Space Algorithm[J]. Laser & Optoelectronics Progress, 2022, 59(12): 1210010

    Download Citation

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

    Category: Image Processing

    Received: Apr. 22, 2021

    Accepted: Jun. 11, 2021

    Published Online: May. 23, 2022

    The Author Email: Yunhong Li (hitliyunhong@163.com)

    DOI:10.3788/LOP202259.1210010

    Topics