FASTTEMPLATE MATCHING BY APPLYINGWINNER-UPDATEON WALSH- HADAMARD DOMAIN

Shou-Der Wei, Shaowei Liu, Shang‐Hong Lai

Abstract

Fasttemplate matching isstrongly demanded formany practical applications related tocomputer vision andimage processing. Inthispaper, we propose a fasttemplate matching method byapplying thewinner-update strategy on theWalsh-Hadamard domain. Bytaking advantage ofthe niceenergypacking property oftheWalsh-Hadamard transformation, wecanjust apply thewinner-update process withasmall numberofWalsh-Hadamard coefficients to reduce thecomputational burden fortemplate matching in animage. Experimental results demonstrate theefficiency androbustness oftheproposed template matching algorithm under different noise levels. IndexTerms-Walsh-Hadmard,
BibTeX
@inproceedings{wei2007fasttemplate,
  author = {Shou-Der Wei and Shaowei Liu and Shang‐Hong Lai},
  title = {{FASTTEMPLATE MATCHING BY APPLYINGWINNER-UPDATEON WALSH- HADAMARD DOMAIN}},
  booktitle = {IEEE International Conference on Acoustics, Speech and Signal Processing},
  year = {2007},
  doi = {10.1109/icassp.2007.366086},
}

Loading related content...

Figure