Evaluation of different features for face recognition in video

PDF

Authors
  1. Gorodnichy, D.O.
  2. Neves, E.
  3. Matwin, S.
  4. Granger, E.
Corporate Authors
Defence Research and Development Canada, Centre for Security Science, Ottawa ON (CAN);OTTAWA UNIV, OTTAWA ON (CAN) SCHOOL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCE;Quebec Univ, Montreal QUE (CAN)
Abstract
With man One of the most critical tasks in automated face recognition technology is the extraction of facial features from a facial images. The most critical task in each face recognition (FR) technology, which contributes the most to the success of particular FR products in particular applications and which is highly protected by industries developing those products, is the extraction of facial features from a facial image. This report presents the performance comparison of several publicly reported feature extraction algorithms for face recognition in video. The evaluated features are Harris corner detection features, FAST (Features from Accelerated Segment Test), GFTT (Good Features To Track), MSER (Maximally Stable Extremal Regions), and HOG (Histograms of Oriented Gradients).
Keywords
video-surveillance;face recognition in video;instant face recognition;watchlist screening;biometrics;reliability;performance evaluation
Report Number
DRDC-RDDC-2014-C250 — Contract Report
Date of publication
01 Sep 2014
Number of Pages
20
DSTKIM No
CA039476
CANDIS No
800512
Format(s):
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: