基于PDE的视频图像数据快速预处理方法

Fast Preprocessing Method of Video Image Data Based on PDE

通过人手识别跟踪模拟实验,研究基于偏微分方程方法(PDE)的视频图像数据快速预处理方法.实验根据颜色模型,利用边缘信息、颜色信息和图像的帧差对人手进行分离,并在边缘检测部分,提出简单背景环境下基于PDE概率统计的一种检测方法,减少了边缘检测的时间,实现了视频图像数据的快速预处理.

A fast preprocessing method of video image data based on partial differential equation method（ PDE） is studied by means of hand identification and tracking simulation experiment. Color model based experiments were carried out to separate the human hand by use of edge information,color information and image frame. On the edge detection part,the detection method based on PDE probability statistics are put forward under a simple background environment,which reduces the time of edge detection to realize the fast preprocessing of video image data.