Pixel Classification of Segmented Objects
Extracting moving objects from video sequences is at the core of various vision applications including video surveillance, content-based video coding etc. Background subtraction is one of the most popular algorithms used for moving object segmentation. The segmentation algorithms often erroneously classify a number of background pixels as moving regions. Two most common misclassifications are shadow and reflections. The goal of this topic is to investigate different post processing methods for object segmentation and develop an efficient pixel classification algorithm suitable for real-time applications.
Requirements: C/C++ programming skills, image processing basic knowledge.
details can be viewed at
http://www.tnt.uni-hannover.de/staff/sho…ssification.pdf
Contact:
Muhammad Shoaib
shoaib@tnt.uni-hannover.de
05117625304