基于背景模型的自动视频分割方法

Automatic Video Segmentation Algorithm Using Background Model

提出一种基于背景模型的自动视频分割方法。该方法结合了背景恢复技术和目标跟踪技术进行视频分割。算法分背景重建、运动目标提取、跟踪技术的使用和后处理4个步骤。同时提出了一种新跟踪算法去除大噪声的干扰。文章阐述了方法的基本思想、理论依据和实现。实验表明:该方法具有很好的效果,具有较强实时性。

An automatic video segmentation algorithm based on background model is proposed in this paper.The proposed algorithm,which combines background restoration technique with tracking technique,consists of four phases:background construction,abstract moving objects,tacking the moving objects and post processing.A new tracking algorithm to remove significance noise is also put forward in this paper.The Algorithm can segment in real-time.The theoretical basis and the algorithm implementation are explained in detail,experiments show the satisfactory results from the algorithm.