• Laser & Optoelectronics Progress
  • Vol. 58, Issue 4, 0410010 (2021)
Yuhao Wang*, Zetian Tang, Minzhe Zhong, Yang Wang, Guangwen Zhao, Caifu Ding, and Chen Yang*
Author Affiliations
  • College of Big Data and Information Engineering, Power Semiconductor Device Reliability Engineering Center of the Ministry of Education, Key Laboratory of Micro-Nano-Electronics and Software Technology of Guizhou Province, Guizhou University, Guiyang, Guizhou 550025, China
  • show less
    DOI: 10.3788/LOP202158.0410010 Cite this Article Set citation alerts
    Yuhao Wang, Zetian Tang, Minzhe Zhong, Yang Wang, Guangwen Zhao, Caifu Ding, Chen Yang. Image Matching Algorithm for Fast Scale-Invariant Feature Transformation Based on Mask Search[J]. Laser & Optoelectronics Progress, 2021, 58(4): 0410010 Copy Citation Text show less
    Flowchart of proposed algorithm
    Fig. 1. Flowchart of proposed algorithm
    Image partition, Mask pyramid, and feature point searching with Mask. (a) Image partition; (b) Mask pyramid; (c) feature point searching with Mask
    Fig. 2. Image partition, Mask pyramid, and feature point searching with Mask. (a) Image partition; (b) Mask pyramid; (c) feature point searching with Mask
    Matching time, correct matching points, and seven-zone circular descriptor. (a) Matching time; (b) correct matching points; (c) seven-zone circular descriptor
    Fig. 3. Matching time, correct matching points, and seven-zone circular descriptor. (a) Matching time; (b) correct matching points; (c) seven-zone circular descriptor
    Image matching results. (a) Brightness transformation; (b) perspective transformation; (c) fuzzy transformation; (d) rotation and scaling transformation; (e) compression transformation
    Fig. 4. Image matching results. (a) Brightness transformation; (b) perspective transformation; (c) fuzzy transformation; (d) rotation and scaling transformation; (e) compression transformation
    Overall time of algorithm, number of feature points, and number of correct matching points. (a) Overall time of algorithm; (b) number of feature points; (c) number of correct matching points
    Fig. 5. Overall time of algorithm, number of feature points, and number of correct matching points. (a) Overall time of algorithm; (b) number of feature points; (c) number of correct matching points
    Descriptor generation time, matching time, and number of correct matching points. (a) Descriptor generation time; (b) matching time; (c) number of correct matching points
    Fig. 6. Descriptor generation time, matching time, and number of correct matching points. (a) Descriptor generation time; (b) matching time; (c) number of correct matching points
    Algorithms for adding extreme category in matching stage. (a) SIFT algorithm; (b) algorithm in Ref. [10]; (c) algorithm in Ref. [11]
    Fig. 7. Algorithms for adding extreme category in matching stage. (a) SIFT algorithm; (b) algorithm in Ref. [10]; (c) algorithm in Ref. [11]
    MethodImageTime for finding feature points /sDescriptor generation time /sMatching time /sTotal time /s
    SIFT1.6724.0121.2057.600
    2.2756.8223.17213.143
    2.2657.1083.17913.373
    2.3357.6045.79614.566
    1.8476.8502.80212.677
    Algorithm inRef. [10]1.5031.7710.5144.453
    2.1073.0061.1296.877
    2.1033.0481.2397.147
    2.2173.2491.4417.626
    1.6752.6311.0816.230
    Algorithm inRef. [11]1.6231.1000.3833.686
    2.2741.8580.8005.575
    2.2341.8022.2027.012
    2.3341.9362.5907.462
    1.8611.6310.7764.822
    Proposed method1.1560.8020.2632.998
    1.6281.3480.4024.132
    1.5491.2280.3484.033
    1.9411.7860.5135.021
    1.5211.4930.4914.297
    Table 1. Running time of different algorithms
    Yuhao Wang, Zetian Tang, Minzhe Zhong, Yang Wang, Guangwen Zhao, Caifu Ding, Chen Yang. Image Matching Algorithm for Fast Scale-Invariant Feature Transformation Based on Mask Search[J]. Laser & Optoelectronics Progress, 2021, 58(4): 0410010
    Download Citation