用于半自动视频对象提取的自适应网格图像分割

Adaptive Mesh Image Segmentation for Semi-Automatic Video Object Extraction

随着 MPEG- 4标准的发展和基于内容的视频处理研究 ,视频对象平面 (VOP)的有效产生成为一个关键问题。本文提出一种基于区域的自适应网格彩色图像分割方法 ,可用于获得半自动视频对象跟踪和提取所需的初始 VOP。该方法利用 CIE L * a* b*色彩空间的特征量 ,对视频序列的第一帧进行三角形网格的分裂与合并。对 MPEG-

With the development of MPEG 4 standard, the efficient generation of video object planes (VOPs) becomes a key issue. This paper proposes a region based color image segmentation scheme using adaptive mesh methods, which can be used to obtain initial VOPs required by semi automatic video object tracking and extraction procedures. This methods utilizes feature values in CIE L*a*b* color space to split and merge triangle meshes of the first frame of the video sequence. Segmentation experiments on MPEG 4 standard test sequences receive good results.