• Laser & Optoelectronics Progress
  • Vol. 56, Issue 5, 051006 (2019)
Xiaoxiao Liu*, Xueliang Ping, and Xinyu Wang
Author Affiliations
  • Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and Technology, School of Mechanical Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China
  • show less
    DOI: 10.3788/LOP56.051006 Cite this Article Set citation alerts
    Xiaoxiao Liu, Xueliang Ping, Xinyu Wang. C-FAST Feature Detection and Matching Algorithm Based on Image Color Information[J]. Laser & Optoelectronics Progress, 2019, 56(5): 051006 Copy Citation Text show less
    Schematic of neighborhood scatter
    Fig. 1. Schematic of neighborhood scatter
    Distribution of feature point via different algrithms. (a) ALOI picture group; (b) FAST; (c) C-FAST
    Fig. 2. Distribution of feature point via different algrithms. (a) ALOI picture group; (b) FAST; (c) C-FAST
    Images from Mikolajczyk
    Fig. 3. Images from Mikolajczyk
    Results of feature matching. (a) C-FAST++, --; (b) FAST++, --; (c) CSIFT++, --; (d) SURF++, --(+ represents clockwise rotation or scale magnification, - represents counterclockwise rotation and scale reduction.)
    Fig. 4. Results of feature matching. (a) C-FAST++, --; (b) FAST++, --; (c) CSIFT++, --; (d) SURF++, --(+ represents clockwise rotation or scale magnification, - represents counterclockwise rotation and scale reduction.)
    Light intensity picture group
    Fig. 5. Light intensity picture group
    Comparison of illumination invariance
    Fig. 6. Comparison of illumination invariance
    Noise performance analysis. C-FAST algorithm in (a) impulse noise environment and (b) Gaussian noise environment; FAST algorithm in (c) impulse noise environment and (d) Gaussian noise environment; CSIFT algorithm in (e) impulse noise environment and (f) Gaussian noise environment; SURF algorithm in (g) impulse noise environment and (h) Gaussian noise environment
    Fig. 7. Noise performance analysis. C-FAST algorithm in (a) impulse noise environment and (b) Gaussian noise environment; FAST algorithm in (c) impulse noise environment and (d) Gaussian noise environment; CSIFT algorithm in (e) impulse noise environment and (f) Gaussian noise environment; SURF algorithm in (g) impulse noise environment and (h) Gaussian noise environment
    Operation results in the natural light environment. (a) C-FAST; (b) FAST; (c) CSIFT; (d) SURF
    Fig. 8. Operation results in the natural light environment. (a) C-FAST; (b) FAST; (c) CSIFT; (d) SURF
    Algorithm matching logarithm
    Fig. 9. Algorithm matching logarithm
    Average time consuming of procedure
    Fig. 10. Average time consuming of procedure
    AlgorithmMatching numberAverage matching rate /%Average time /s
    M1M2N1N2
    C-FAST15316512412650.910.031470
    FAST23534414716338.080.026257
    CSIFT3261109213053.060.051642
    SURF1271521708248.230.037182
    Table 1. Parameter comparison of different algorithms
    Xiaoxiao Liu, Xueliang Ping, Xinyu Wang. C-FAST Feature Detection and Matching Algorithm Based on Image Color Information[J]. Laser & Optoelectronics Progress, 2019, 56(5): 051006
    Download Citation