TY - GEN A1 - Steć, Piotr PB - Zielona Góra: Oficyna Uniwersytetu Zielonogórskiego N2 - This thesis is related to the problem of segmentation of video sequences. Automatic video segmentation plays important role in many applications such as: object oriented coding of digital video, remote surveillance, interactive multimedia presentations and video editing. The main problem considered in this work is ability of segmentation sequences acquired with a moving camera which parameters are unknown. As a main analytical tool the Fast Marching Method was used as an active contour propagation method. Since it is very general algorithm, it was necessary to adopt it to the problem that was considered in this thesis. N2 - As a result of research two related algorithms of video segmentation were developed, however, each of them fulfill different assumptions and perform segmentation in a different way. The first algorithm separates foreground objects from a background in two steps and can be applied to a wide class of video sequences. The first step of the algorithm uses Fast Marching Method to perform initial segmentation while the second step improves segmentation quality. The second algorithm developed in this thesis is able to segment a video sequence into many separate objects but is limited to the sequences containing only rigid objects that perform only translational motion. In the latter case, the multilabel version of the Fast Marching Method was used. L1 - http://www.zbc.uz.zgora.pl/Content/2286/HTML/book.pdf L2 - http://www.zbc.uz.zgora.pl/Content/2286 KW - segmentacja KW - sekwencje wizyjne KW - kompresja KW - algorytm szybkiej propagacji KW - propagacja KW - rozprawa doktorska KW - telekomunikacja KW - nauki techniczne KW - przetwarzanie obrazów T1 - Segmentation of Colour Video Sequences Using the Fast Marching Method UR - http://www.zbc.uz.zgora.pl/dlibra/docmetadata?id=2286 ER -