期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
基于PAF的深度图人体姿态估计 被引量:1
1
作者 刘涛 杨璐 邵肖伟 《智能计算机与应用》 2020年第1期103-108,共6页
采用Part Affinity Field(PAF部分关联域)与卷积神经网络(CNN)结合的模型,解决深度图像下人体姿态估计问题。首先,通过CNN得到人体的一组特征图。然后,使用CNN分别提取其关节点信息以及PAF信息。最后,采用图论的匹配方法对各个关节点进... 采用Part Affinity Field(PAF部分关联域)与卷积神经网络(CNN)结合的模型,解决深度图像下人体姿态估计问题。首先,通过CNN得到人体的一组特征图。然后,使用CNN分别提取其关节点信息以及PAF信息。最后,采用图论的匹配方法对各个关节点进行推理,将同一个人的关节点连接起来得到估计结果。实验结果表明,文中方法可以很好应用于深度图场景下。 展开更多
关键词 部分关联域 卷积神经网络 深度图 人体姿态估计 图论匹配
下载PDF
Induced Matching Number of the Plane Grid Graph
2
作者 LUXiao-xu PEIMing +1 位作者 YAOWei-li ZHOUJu 《Chinese Quarterly Journal of Mathematics》 CSCD 2004年第1期95-100,共6页
An induced matching M in a graph G is a matching such that V(M) induces a 1-regular subgraph of G. The induced matching number of a graph G, denoted by I M(G), is the maximum number r such that G has an induced matchi... An induced matching M in a graph G is a matching such that V(M) induces a 1-regular subgraph of G. The induced matching number of a graph G, denoted by I M(G), is the maximum number r such that G has an induced matching of r edges. Induced matching number of Pm×Pn is investigated in this paper. The main results are as follows:(1) If at least one of m and n is even, then IM(Pm×Pn=[(mn)/4].(2) If m is odd, then 展开更多
关键词 induced matching induced matching number α-set
下载PDF
Measuring Propeller Blade Width Using Binocular Stereo Vision 被引量:3
3
作者 张建德 陆金桂 《Journal of Marine Science and Application》 2011年第2期246-251,共6页
Propeller blade width measurement has been extensively studied in the past using direct and indirect methods, and it plays a great role in determining the quality of the finished products. It has surveyed that previou... Propeller blade width measurement has been extensively studied in the past using direct and indirect methods, and it plays a great role in determining the quality of the finished products. It has surveyed that previous techniques are usually time-consuming and erroneous due to a large number of points to be processed in blade width measurement. This paper proposes a new method of measuring blade width using two images acquired from different viewpoints of the same blade. And a new feature points matching approach for propeller blade image is proposed in stereo vision measurement. Based on these, pixel coordinates of contour points of the blade in two images are extracted and converted to real world coordinates by image algorithm and binocular stereo machine vision theory. Then, from the real world coordinates, the blade width at any position can be determined by simple geometrical method. 展开更多
关键词 blade width binocular stereo vision propeller: vision measurement
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部