Foreground region detection and tracking for fixed cameras

Türdü, Deniz (2010) Foreground region detection and tracking for fixed cameras. [Thesis]

[thumbnail of DenizTurdu.pdf] PDF

Download (783kB)


For real-time foreground detection on videos, probabilistic modeling for background and foreground colors are widely used. Stauffer and Grimson's model is very successful for foreground segmentation. In this method, each pixel is modeled independently with Gaussian mixtures. Explicit foreground probabilities for pixels are not calculated. Spatial and temporal continuity of pixels are omitted. In this thesis, we obtain foreground probabilities for the pixels using Stauffer and Grimson's model and apply hysteresis thresholding to utilize spatial continuity of pixels. For the same purpose, we also use Markov Random Field modeling and optimizations. To leverage the temporal continuity of pixels, mean-shift tracking is integrated into the segmentation to increase accuracy. Wherever applicable, we combine some of these improvements together. Our work shows that using the probabilistic approach with different enhancements results in much higher segmentation accuracy.
Item Type: Thesis
Uncontrolled Keywords: Background modeling. -- Object detection. -- Foreground segmentation. -- Tracking. -- Gaussian mixture model. -- Image enhancement. -- Image segmentation. -- Image modelling. -- Markov model. -- Markov random fields. -- Object recognition. -- Object tracking. -- Arkaplan modelleme. -- Nesne tanıma. -- Önplan çıkarımı. -- Nesne takibi. -- Gauss karışım modeli. -- Görüntü iyileştirme. -- Görüntü bölütleme. -- Görüntü modelleme. -- Markov modeli. -- Markov rastgele alanları. -- Nesne izleme.
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7800-8360 Electronics
Divisions: Faculty of Engineering and Natural Sciences > Academic programs > Electronics
Faculty of Engineering and Natural Sciences
Depositing User: IC-Cataloging
Date Deposited: 19 Apr 2012 16:29
Last Modified: 26 Apr 2022 09:55

Actions (login required)

View Item
View Item