Evaluation of the most appropriate Kernel function for the purpose of feature extraction in face recognition in video surveillance systems

Sepehr Damavandinejadmonfared*, Sina Ashooritootkaboni, Taha Bahraminezhad Jooneghani

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionpeer-review

Abstract

It is believed that Kernel Principal Component Analysis (KPCA) merits the performance of Principal Component Analysis big time. KPCA is the nonlinear extend of Principal Component Analysis (PCA) meaning that in KPCA the input data is first mapped by using nonlinear Kernel and then PCA is performed on the mapped data. Kernel function has been categorized into four different categories such as Linear, Polynomial, Gaussian, and Sigmoid. In this work the performance of the mentioned methods on Surveillance Camera Faces database is observed to determine the significance of Polynomial KPCA compared to other types of that in terms of video surveillance recognition.

Original languageEnglish
Title of host publicationProceedings of the 2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012
Pages1131-1134
Number of pages4
Volume2
Publication statusPublished - 2012
Externally publishedYes
Event2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012 - Las Vegas, NV, United States
Duration: 16 Jul 201219 Jul 2012

Other

Other2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012
CountryUnited States
CityLas Vegas, NV
Period16/07/1219/07/12

Keywords

  • Biometrics
  • Face recognition
  • Kernel Principal Component Analysis (KPCA)
  • Pattern recognition
  • Video surveillance systems

Fingerprint Dive into the research topics of 'Evaluation of the most appropriate Kernel function for the purpose of feature extraction in face recognition in video surveillance systems'. Together they form a unique fingerprint.

Cite this