Open AccessStabilising illumination variations in motion detection for surveillance applications
Author(s) -
Vujović Igor,
Šoda Joško,
Kuzmanić Ivica
Publication year - 2013
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/iet-ipr.2013.0169
Subject(s) - wavelet , artificial intelligence , false alarm , constant false alarm rate , computer science , wavelet transform , computer vision , energy (signal processing) , pattern recognition (psychology) , motion detection , motion (physics) , mathematics , statistics
Since illumination variations may cause the misinterpretation of data for various higher vision applications and algorithms, this study aims to reduce such influence. In order to obtain a motion mask, which is input for a higher vision application, wavelet coefficients are calculated by applying two‐dimensional lifting wavelet transform with two mother wavelets. Energy is calculated from the obtained wavelet coefficients. Morphological operations are used to improve output image. The developed algorithm is a robust algorithm further reducing false alarm readings caused by illumination variations (better false detection rate and percentage of correct classifications).