Laser & Optoelectronics Progress, Volume. 57, Issue 2, 21005(2020)

Image Registration Based on Spectral and Spatial Feature Matching

Chen Zefeng1, Wu Qingyang1,2、*, Chen Shunzhi1, Li Qifeng1, Lu Xiaoting1, and Huang Haotao1
Author Affiliations
  • 1College of Electronic Science and Technology, Shenzhen University, Shenzhen, Guangdong 518060, China
  • 2College of Big Data and Internet, Shenzhen Technology University, Shenzhen, Guangdong 518118, China
  • show less
    Figures & Tables(10)
    Spectrograms of two images. (a)(b) Spectrograms in polar coordinate system; (c)(d) spectrograms in rectangular coordinate syatem
    Two images to be registered with only relative translation
    Diagram of improved mesh matching algorithm
    Relative position of two images to be registered when value of function S is the smallest
    Registered images of two images with inconspicuous features. (a) Gray difference mean value of overlapped area is 6.4564×10-4; (b) gray difference mean value of overlapped area is 8.0699×10-4
    Two images to be registered
    Spectrograms of two images. (a)(b) Spectrograms in polar coordinate system; (c)(d) spectrograms in rectangular coordinate system
    Similarity curve of two spectrograms in rectangular coordinate system obtained by SAD
    Gray level difference curves obtained by coarse matching and fine matching of SAD. (a) Coarse matching; (b) fine matching
    Images after coordinate transformation. (a) Image after registration of two images; (b) image after continuous mosaic of multiple images
    Tools

    Get Citation

    Copy Citation Text

    Chen Zefeng, Wu Qingyang, Chen Shunzhi, Li Qifeng, Lu Xiaoting, Huang Haotao. Image Registration Based on Spectral and Spatial Feature Matching[J]. Laser & Optoelectronics Progress, 2020, 57(2): 21005

    Download Citation

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

    Category: Image Processing

    Received: May. 5, 2019

    Accepted: --

    Published Online: Jan. 3, 2020

    The Author Email: Qingyang Wu (wuqingyang@sztu.edu.cn)

    DOI:10.3788/LOP57.021005

    Topics