Semi-automatic semantic video object extraction by active contour model

摘要

An algorithm to extract the semantic video object in the first frame of a video sequence is proposed. The extracted closed object contour provides an accurate video object model for automatic object extraction in the following frames. Two polygons are input by the user to specify the area in which the object boundary is located. Based on this, edges which belong to the video object of interest are selected by a local object contour model. Then an active contour model (snake) is applied to refine the closed object contour.