A Method for Generating Face Video Light and Shadow Migration Based on a Single Reference Video
A video, light and shadow technology, applied in the field of facial expression video light migration, can solve the problem of not taking into account the smooth transition relationship between the video reference frame and the target frame expression difference, so as to avoid the phenomenon of light and shadow jitter, reduce technical difficulty, and save costs Effect
Active Publication Date: 2015-08-12
BEIHANG UNIV
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The above image-based methods do not take into account the smooth transition relationship between videos and the expression difference between the reference frame and the target frame.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0033] The present invention will be further described in detail below in conjunction with the accompanying drawings, so that those skilled in the art can implement it with reference to the description.
[0034] In the reference video used by the present invention, the light and shadow changes of the face, the expression and the posture remain unchanged. The target videos were captured under approximately uniform lighting conditions with rich facial expressions. The number of frames of the reference video and the target video are the same, and the frames of the reference video correspond to the frames of the target video one-to-one.
[0035] Such as figure 1 , 2 Shown in and 4, the present invention comprises following 6 basic steps:
[0036] Step 1, select the same reference video with the time length of the target video, select a frame every n frames in the target video as the target key frame, and select a frame every n frames in the reference video as the reference key ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention belongs to the virtual reality and computer visual field, and discloses a human face video light and shadow migration generation method based on a single reference video. A front human face video (a target video) under approximately uniform illumination and a human face video (the reference video) under variable illumination are given, and light and shadow effects of the face portion of the reference video are migrated to the target video to generate a realistic video which is similar to the illumination effect of the reference video. The human face video light and shadow migration generation method based on the single reference video includes illumination information migration of target key frames and illumination information dissemination of target non-key frames, wherein the illumination information migration of the target key frames includes aligning of the target key frames and target reference key frames, smoothness maintenance of edges, extraction of illumination information in a skin region, and interpolation of illumination information in a non-skin region. The human face video light and shadow migration generation method based on the single reference video can be widely used and popularized in video post-production, personal video editing and other application scenarios.
Description
technical field [0001] The invention relates to the field of virtual reality and computer vision, in particular to a method for light migration of human facial expression video based on a single reference video. Background technique [0002] Virtual reality fusion technology based on video material is an important part of virtual reality technology. Since the virtual scene and the scene objects may come from different materials, the lighting effects of the scene objects and the scene may be quite different. However, the virtual-real fusion scene requires the scene and the scene objects to have consistent lighting effects. However, in the process of scene object acquisition, the lighting conditions of the scene cannot be predicted, and the lighting conditions of the acquisition environment cannot be set according to the scene lighting conditions. A feasible solution is to capture scene object videos under uniform lighting conditions, and then transfer the required lighting c...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T5/50
Inventor 陈小武吴洪宇赵沁平杨梦夏方志宏
Owner BEIHANG UNIV