Inspection System based on Computer Vision

Authors

  • M. Petkevicius Kaunas University of Technology
  • A. Vegys Kaunas University of Technology
  • T. Proscevicius Kaunas University of Technology
  • A. Lipnickas Kaunas University of Technology

DOI:

https://doi.org/10.5755/j01.eee.116.10.889

Abstract

State-of-the-art inspection systems are often used for surveillance purposes, which inspect human behaviour or detect suspicious activities using visual data. Crowd detection and individual person tracking in real-time require robust human detection. We propose a hybrid method, which is based on a combination of histogram of oriented gradients (HoG) and Haar-like features. The former is used for human silhouette detection and the latter for face detection. Novel method ensures more precise human detection and the ability to estimate looking direction. Our method shows better results then HoG or Haar-like features independently. Ill. 7, bibl. 8 (in English; abstracts in English and Lithuanian).

http://dx.doi.org/10.5755/j01.eee.116.10.889

Downloads

Published

2011-12-06

How to Cite

Petkevicius, M., Vegys, A., Proscevicius, T., & Lipnickas, A. (2011). Inspection System based on Computer Vision. Elektronika Ir Elektrotechnika, 116(10), 81-84. https://doi.org/10.5755/j01.eee.116.10.889

Issue

Section

SYSTEM ENGINEERING, COMPUTER TECHNOLOGY