| Research Abstract |
The purpose of this study is to extract the subtle nuance of handwritten-animated motions automatically by fully utilizing the motion captured data. I considered a method of calculating motions of omitted and exaggerated expressions using a computer, by altering and modifying not only the timing but also the posing of the character in the subjected frames, from which only the useful frames were extracted as keyframes from the motion capture. This thesis consists of 6 stages. Stage 1 is an introduction that describes the background, purpose and meaning of the study.
In stage 2, I extracted some of the important keys of drawings from the sampled continuous frames and discussed the result and evaluated the difference between the motion of the motion capturing and the extracted keyframe motion that was spline interpolated. As a result, we were able to extract keyframes accurately.
In stage 3, I discussed the treatment process for the poses of the extracted keyframes in stage 2. Expression de
… More
forming was done for hand-drawn animation and this expression deforming creates characteristic motions. In order to create these characteristic motions, we analyzed the hand drawings to identify the factor. Based on these facts, we considered the process to grasp poses by quantity automatically.
In stage 4, I discussed deforming time that coincided with limited animation, rearranged the whole frame by altering the timing of each keyframe and frame. To produce effective motions in a limited number of drawings, the continuous motion needs to be imbalanced instead of being equal. Then by adjusting the smallest set value of keys of before and behind the keyframe to show proximity or distance, we were able to express the possibility of making high and low speed in timing.
In stage 5, we experimented with evaluating impressions by producing trial animation images to verify the effectiveness of the method. The participants were 30 university students, and for the experiment conditions, we repeated the trial animation image and the hand-drawn animation image 10 times each on a computer monitor. We used the SD method for measurement, and used 25 words that were used in the past. Then we drew the curved line from the mean value and verified the psychological approximation of trial animation and hand-drawn animation. As a result, the impressions of the two methods were about the same.
In stage 6, we indicated the possibility of producing high quality images that can be brought into practical use, and with the outcome of this study, we proved the effectiveness of the limited animation production method using motion capturing. Finally, I discussed the future prospective and some problems remaining as a conclusion. Less
|
