中文摘要：

在传统时空联合算法的基础上,提出了一种基于定时段区域补偿的视频对象分割后处理算法。首先,通过对帧差图像进行噪声抑制和膨胀连接获得变化检测模板;然后,对原始图像进行开闭重构简化,求取形态学梯度,通过对形态学梯度图像进行非线性变换和梯度等级划分并最终由分水岭算法获得对象的精确边界,通过比例运算提取出视频对象的初始二值化模板;最后,通过定时段区域补偿获得最终的完整视频对象模板。实验结果证明了该算法的正确性和有效性。

英文摘要：

A post-processing algorithm of video object segmentation based on region compensation in fixed time interval is proposed on the base of the traditional spatial-temporal video object segmentation algorithm.Firstly,an image variation detection template is gotten by noise suppression and dilatation linking of the image differences of sequential frames;Secondly,the morphological gradient is calculated by the simplification of morphological opening and closing reconstruction of the original image.The accurate bo...