VOGUE Tutorial
This tutorial will guide you through the main functionalities
of VOGUE as a tool for Video Object Generation. For that you will need the screen cam player of lotus. Just go along reading and
click on the links. This will show a demonstration on how to execute the
described functionality.
Introduction
Three main automatic and semi-automatic tools are provided
by VOGUE. With the spatial segmentation module, you can interactively
segment still images. Once a mask corresponding to the object of interest
has been created for a given image, it can be tracked throughout the video
sequence using the object tracking module. When you are mainly interested
in moving objects, the temporal segmentation module automatically
detects them. Although an initial mask is not needed to carry out temporal
segmentation, its use is optional and can improve the results. The automatic
execution of both tracking and temporal segmentation algorithms may be
stopped at any given time for interactive correction.
Tutorial
-
Load a video sequence.
-
Spatial segmentation.
-
STEP 1: Produce an initial partition. This can be done in
two ways Introducing markers
or selecting a number of regions.
-
STEP 2: Re-segment or merge regions.
-
STEP 3: Contour adjustment.
-
STEP 4: Manual refinement.
-
STEP 5: Set background.
-
Tracking the initial mask.
-
Setting parameters.
-
Start the automatic execution.
-
Modifications during execution.
-
Temporal segmentation.
-
Setting parameters.
-
Start the automatic execution.
-
Modifications during execution.
-
Save the results.