• Acta Optica Sinica
  • Vol. 37, Issue 11, 1115003 (2017)
Guofeng Tong, Yong Li, Nan Liu*, and Guangxu Ji
Author Affiliations
  • College of Information Science and Engineering, Northeastern University, Shenyang, Liaoning 110819, China
  • show less
    DOI: 10.3788/AOS201737.1115003 Cite this Article Set citation alerts
    Guofeng Tong, Yong Li, Nan Liu, Guangxu Ji. Mixed Feature Extraction and Matching for Large Affine Scene[J]. Acta Optica Sinica, 2017, 37(11): 1115003 Copy Citation Text show less
    Flow chart of mixed feature extraction and matching algorithm
    Fig. 1. Flow chart of mixed feature extraction and matching algorithm
    ASIFT feature extraction of aerial image. (a) Original image; (b) ASIFT features
    Fig. 2. ASIFT feature extraction of aerial image. (a) Original image; (b) ASIFT features
    MSER feature extraction of aerial image. (a) Original image; (b) MSER features; (c) improved MSER features
    Fig. 3. MSER feature extraction of aerial image. (a) Original image; (b) MSER features; (c) improved MSER features
    Contrast of extreme value regions before and after simulated street view image screening. (a) Original image; (b) MSER features; (c) improved MSER features
    Fig. 4. Contrast of extreme value regions before and after simulated street view image screening. (a) Original image; (b) MSER features; (c) improved MSER features
    Contrast of extreme value regions before and after aerial image screening. (a) Original image; (b) MSER features; (c) improved MSER features
    Fig. 5. Contrast of extreme value regions before and after aerial image screening. (a) Original image; (b) MSER features; (c) improved MSER features
    Feature extraction and matching result of multi-feature fusion
    Fig. 6. Feature extraction and matching result of multi-feature fusion
    Aerial images and simulated street images of Northeastern University
    Fig. 7. Aerial images and simulated street images of Northeastern University
    Contrast of correct matching numbers of large affine image with different algorithms
    Fig. 8. Contrast of correct matching numbers of large affine image with different algorithms
    Contrast of correct matching rates of large affine image with different algorithms
    Fig. 9. Contrast of correct matching rates of large affine image with different algorithms
    Dense point cloud model of Northeastern University library joint space-ground modeling
    Fig. 10. Dense point cloud model of Northeastern University library joint space-ground modeling
    Image numberSerial running time /sParallel running time /sAcceleration ratio /%
    A_01&S_01224.5514.7815.19
    A_02&S_02229.5315.2115.09
    A_03&S_05226.0814.9315.14
    A_04&S_01226.3515.6014.51
    A_05&S_04231.7015.9114.56
    A_06&S_04227.3714.0016.24
    A_08&S_08226.2115.1914.89
    A_08&S_09226.8714.8115.32
    A_10&S_10231.1514.0516.45
    Table 1. Contrast of running time statistics of serial-parallel feature extraction
    Guofeng Tong, Yong Li, Nan Liu, Guangxu Ji. Mixed Feature Extraction and Matching for Large Affine Scene[J]. Acta Optica Sinica, 2017, 37(11): 1115003
    Download Citation