Summary: | 碩士 === 銘傳大學 === 資訊傳播工程學系碩士班 === 98 === Using the camera to photograph and record daily life has become a trend in recent years. To understand the long term action process, one needs to watch video or line up the pictures. Besides, it is difficult to see at a glance the magnificent landscape. To solve those inconvenient, this study uses the modified temporal difference method to extract moving objects, then the background reconstruction process can create an outline of the image contains objects. For the case of the camera moving, this study combines the global and local motion information, determines the motion model of the camera, and then makes a dynamic panoramic image. The background and moving objects can be segmented that will be created as an images summary of object activities. If the video content is a landscape view, the method will use the full motion information to establish a dynamic panoramic image.
|