基于自适应噪声滤波的视频对象分割算法

Video Object Segmentation Algorithm Based on Adaptive Noise Filtering

介绍了基本的视频运动对象分割算法,针对变化区域检测算法中固定阈值滤噪的不足,提出了自适应噪声滤波的方法,同时利用计算机图形学中的扫描线填充思想得到变化区域的具体位置,实现视频对象的分割。实验证明了算法的有效性。

The basic algorithm of video object segmentation is introduced. In consideration of the shortcoming of the fixed threshold noise filtering, an adaptive noise filtering algorithm is proposed. The exact position of changing region is acquired by using scan-line filling method, and then video object segmentation is completed. Experimental results show that the proposed algorithm is effective.