MOBILE OBJECT VIDEO IMAGE COMPOSTING METHOD, ITS DEVICE AND RECORDING MEDIUM RECORDED WITH THE METHOD

摘要

PROBLEM TO BE SOLVED: To provide a mobile object video image compositing method and its device with which mobile object information is extracted from video images in a form to be composited with the memory of a small capacity and new video images are composited based thereon. SOLUTION: First, a mobile object segmentation means 1 segments a 1st frame contour line from a 1st frame of an inputted image. A characteristic point arrangement means 2 places a 1st frame characteristic point on the contour line and the inside thereof. Furthermore, the mobile object segmentation means 1 segments 2nd-n-th frame contour lines. A characteristic point tracing means 3 compares the 1st frame with the 2nd frame and obtains a corresponding position on the 2nd frame of the 1st frame characteristic point to use it as a 2nd frame characteristic point. Then a means 3 compares the 2nd frame with the 3rd frame and obtains a corresponding position of the 2nd frame characteristic point on the 3rd frame and to use it as a 3rd frame characteristic point. Similarly, the means 3 obtains the n-th frame characteristic point. A video image composite information storage means 4 stores two images of the 1st frame and the n-th frame and the 1st-n-th frame characteristic points. A video image composting means 5 composites the video images by means of warps and morphing or the like to match with the movements of the characteristic points from the stored video composite information.