An automatic and accurate moving object extraction algorithm is proposed to segment head-shoulder video sequence.
针对头肩视频序列，本文提出了一种头肩视频对象的自动精确提取算法。
Experimental results show that the proposed object extraction algorithm and shape descriptor are of high accuracy and reliability. They outperform their corresponding alternatives in literature.
实验结果表明本文提出的物体提取算法和形状描绘子准确性较高，可靠性较强，并优于现有的同类方法。
The algorithm has four major steps, feature extraction, object identification, auxiliary image creation and object mining.
算法主要由四步组成：属性抽取，对象鉴别，辅助图像创建和对象挖掘。
A fast recognition algorithm is presented for interesting region extraction of an object based on corner features.
本文给出了一种基于角特征的目标兴趣区的识别算法。
Using color image that makes extraction and identification of object easily, at the same time, it also simplifies algorithm and reduces...
这样就能够更好地对一些具有颜色特征的目标进行提取与识别，简化算法，缩短运算时间。
Using color image that makes extraction and identification of object easily, at the same time, it also simplifies algorithm and reduces...
这样就能够更好地对一些具有颜色特征的目标进行提取与识别，简化算法，缩短运算时间。
应用推荐