The Extraction Of Element Factor In Storyboard Of Animation Video Sequences

In the cartoon making process, directors first produces a storyboard - a series of drawn images that graphically portray the actions describe d in the script. It is at this stage that the major decisions are taken as far as the film's content is concerned, such as composition, positions of characters, shot length and camera movements. All of these factors constitute the style of the director in the produced animations.Currently, to learn the style of the director relies on mainly viewing relevant animations, the number of animation products is however huge, and it is time consuming to view large number of animation products because viewers have to go through every frame in the animations. We attempt to extract some major elements from existing animation sequences and provide them to the viewers. Another possible application of our work is to provide those factors for the ieference of the director in the new animation series, thus we extend computer assisted animation to computer assisted storyboard making.We developed a system aiming at achieving the goal mentioned above. The system includes four modules: shot segmentation based histogram, identification of camera movement using the optical flow technique, recognition of different type of shots as well as the number of characters in the scene based on face detection. The input of the system is cartoon sequences, the system first detects the shots contained in the sequence, and then process each shot to obtain the information of camera movements, shot type and the number of characters. We demonstrate our system with experiment results and discuss them in the thesis.