Édition de Video Basée-exemple
Oriel Frigo · HAL (Le Centre pour la Communication Scientifique Directe) · 2016
The objective of this thesis is to provide new techniques for example-based video editing. We address three related problems: color transfer, tonal stabilization and style transfer. The first problem consists in transferring colors from anexample image. We provide an efficient solution that maps color distributions by optimal transportation and performs a smooth color transformation that creates no artifacts.The second problem is related to tonal fluctuation in videos. Due to the automatic settings of consumer cameras, the colors of objects in image sequences might change over time. We present a fast and computationally light method to stabilize video tonal appearance using a minimally-viable color correction model.Finally, we discuss the problem of transferring the style of an example image to a source image. The complex notion of image style is considered as a local texture transfer, eventually coupled with a global color transfer. For the local texturetransfer, we propose a new patch-based method using an adaptive partition that captures the style of the example image and preserves the structure of the source image.