Optics and Precision Engineering, Volume. 32, Issue 6, 915(2024)

3D reconstruction technique based on SURF-OKG feature matching

Lei ZHANG1... Yan SHI1,*, Wenyong LU1, Rui XU1, Zhan JIN2, Weijie LUO3, Yi CEHN1, Chunliu ZHAO1 and Chunlian ZHAN1 |Show fewer author(s)
Author Affiliations
  • 1College of Optical and Electronic Technology, China Jiliang University, Hangzhou3008, China
  • 2Zhejiang Visual Intelligence Innovation Center Co., Ltd, Hangzhou31115, China
  • 3Zhejiang Peking University Institute of Information Technology Advanced Research, Hangzhou11215, China
  • show less
    Figures & Tables(25)
    Flowchart of SURF algorithm for feature matching
    Adaptive median filter
    Scale space extremum detection
    Result of feature point matching using the SURF algorithm
    SURF-OKG matching algorithm flow chart
    FAST corner detection effect
    Illustration of a K-D tree
    Matching results of the SURF-OKG algorithm
    Matching situation corresponding to different thresholds
    Algorithm comparison chart
    Correct matching rate of different algorithms
    Matching results for the image labeled Bikes
    Matching results for the image labeled Bark
    Matching results for the image labeled Leuven
    Analysis of matching results for Mikolajcyzk dataset images
    Epipolar correction result
    System physical map
    3D reconstruction results
    3D reconstruction results of letters
    3D reconstruction results of step blocks
    Analysis of measurement results
    • Table 1. Preliminary feature matching results of the SURF algorithm

      View table
      View in Article

      Table 1. Preliminary feature matching results of the SURF algorithm

      匹配数正确匹配数正确匹配率/%
      传统SURF714969.01
      改进SURF755776.65
    • Table 2. Secondary feature matching results of the SURF-OKG algorithm

      View table
      View in Article

      Table 2. Secondary feature matching results of the SURF-OKG algorithm

      匹配数正确匹配数正确匹配率/%
      本文算法20218591.58
    • Table 3. Feature point matching results of different algorithms

      View table
      View in Article

      Table 3. Feature point matching results of different algorithms

      匹配数正确匹配数正确匹配率/%耗时/s
      SIFT45438885.264.023
      SURF18214278.420.748
      ORB49836473.271.631
      SURF-OKG48544892.472.891
    • Table 4. Binocular camera calibration results

      View table
      View in Article

      Table 4. Binocular camera calibration results

      内参矩阵

      ML=3.581 121 46×10301.259 401 42×10303.577 615 47×1031.032 672 79×103001

      MR=3.595 605 92e+0301.194 383 20×10303.597 358 98×1031.010 168 02×103001

      外参矩阵

      DL=-3.508 331 74×10-22.568 322 66×10-1-4.233 416 33×10-4              5.586 009 70×10-5-1.239 371 92×10-1]

      DR=-3.219 153 46×10-22.243 644 92×10-1-2.211 489 98×10-4              -8.376 793 01×10-4-1.895 272 40×10-1]

      旋转矩阵R=9.323 543 05×10-1-4.057 619 26×10-33.615 231 47×10-18.871 245 94×10-49.999 596 85×10-18.935 364 92×10-3-3.615 448 28×10-1-8.010 209 87×10-39.323 203 17×10-1
      平移矩阵DL=-2.637 902 1×1021.013 302 0×10-16.264 246 6×10
    Tools

    Get Citation

    Copy Citation Text

    Lei ZHANG, Yan SHI, Wenyong LU, Rui XU, Zhan JIN, Weijie LUO, Yi CEHN, Chunliu ZHAO, Chunlian ZHAN. 3D reconstruction technique based on SURF-OKG feature matching[J]. Optics and Precision Engineering, 2024, 32(6): 915

    Download Citation

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

    Category:

    Received: Aug. 21, 2023

    Accepted: --

    Published Online: Apr. 19, 2024

    The Author Email:

    DOI:10.37188/OPE.20243206.0915

    Topics