• Laser & Optoelectronics Progress
  • Vol. 62, Issue 10, 1037002 (2025)
Lang Liu1, Yanlin Shao1,*, Qihong Zeng2, Kunpeng Zhao1..., Changhui Zhou1, Peijin Li1 and Rui Zeng1|Show fewer author(s)
Author Affiliations
  • 1School of Geosciences, Yangtze University, Wuhan 430100, Hubei , China
  • 2Research Institute of Petroleum Exploration & Development, Beijing 100000, China
  • show less
    DOI: 10.3788/LOP241999 Cite this Article Set citation alerts
    Lang Liu, Yanlin Shao, Qihong Zeng, Kunpeng Zhao, Changhui Zhou, Peijin Li, Rui Zeng. Lithology Segmentation Method with Efficient Channel Attention Based on Multiple Eigenvalues of Outcrop Voxel[J]. Laser & Optoelectronics Progress, 2025, 62(10): 1037002 Copy Citation Text show less
    Structure of MGECA network
    Fig. 1. Structure of MGECA network
    Structure of multi-granularity convolutional neural network
    Fig. 2. Structure of multi-granularity convolutional neural network
    Structure of dual-channel group convolution module
    Fig. 3. Structure of dual-channel group convolution module
    Structure of ECA-Transformer encoder
    Fig. 4. Structure of ECA-Transformer encoder
    Camera photograph and calibrated laser scan result of outcrop in Crescent Bay
    Fig. 5. Camera photograph and calibrated laser scan result of outcrop in Crescent Bay
    Schematic diagram of optimal parameter separability
    Fig. 6. Schematic diagram of optimal parameter separability
    Comparison results of ablation experiments
    Fig. 7. Comparison results of ablation experiments
    Comparison of identification results for laser outcrop dataset in Crescent Bay
    Fig. 8. Comparison of identification results for laser outcrop dataset in Crescent Bay
    BaselineMulti-granularity convolutionDual-channel group convolutionECAOAmAccmIoU
    84.180.258.8
    86.281.663.5
    88.883.868.2
    90.685.770.4
    Table 1. Ablation experimental results of MGECA
    MethodOAmAccmIoU
    SVM56.750.537.5
    DGPoint58.953.445.7
    DGCNN85.172.456.1
    DBDA88.078.463.5
    MGECA90.685.770.4
    Table 2. Results of comparative experiments on laser outcrop datasets in Crescent Bay
    MethodLithologyAA /%Every epoch training time /s
    SVMSandstone26.775
    Conglomerate40.1
    Mudstone57.2
    Grass62.1
    DGPointSandstone35.9164
    Conglomerate79.8
    Mudstone68.4
    Grass73.1
    DGCNNSandstone11.2235
    Conglomerate94.1
    Mudstone65.6
    Grass78.4
    DBDASandstone67.2110
    Conglomerate84.6
    Mudstone70.9
    Grass89.4
    MGECASandstone74.3131
    Conglomerate91.6
    Mudstone89.2
    Grass58.6
    Table 3. Identification accuracy of single target type by different methods
    Lang Liu, Yanlin Shao, Qihong Zeng, Kunpeng Zhao, Changhui Zhou, Peijin Li, Rui Zeng. Lithology Segmentation Method with Efficient Channel Attention Based on Multiple Eigenvalues of Outcrop Voxel[J]. Laser & Optoelectronics Progress, 2025, 62(10): 1037002
    Download Citation