基于点云模型的人体尺寸自动提取方法

来源期刊:中南大学学报(自然科学版)2014年第8期

论文作者:赖军 王博 付全 吴壮志

文章页码:2676 - 2684

关键词:人体测量学;模型分割;测点识别;人体尺寸

Key words:anthropometry; model segmentation; landmark recognition; body size

摘    要:基于点云数据的人体尺寸自动提取是三维人体非接触测量的难点问题,提出一种基于点云数据的人体尺寸自动提取方法。其步骤为:对人体点云去噪并建立标准测量坐标系;给出一种自动分割算法将人体分为6部分;将人体测点分为极值点、局部极限点和一般点3类,在模型分割的基础上给出每类测点的识别算法;最后给出人体测量项目的计算方法。研究结果表明:与传统方法相比,本方法具有全自动、速度快、操作便捷的特点;所提出的测点识别算法和测量项目的计算方法是有效的,在准确率上满足GB/T 23698—2009的要求。

Abstract: Considering that body size extraction based on point cloud data is a difficult problem for three-dimensional non-contact measurement of the human body, an automatic extraction method of human sizes was presented based on human point cloud data. The procedures were as follows. Firstly, the point cloud data were de-noisesed and a standard coordinate was built for measurement. Secondly, a method to segment the human body into six parts was proposed so that the landmarks could be recognized more efficiently and accurately. Thirdly, the recognition methods of human landmarks were introduced which were divided into three classes, i.e. the global extrema,the local extrema and the points that haven’t obvious characteristics. Finally, the human sizes were calculated based on the identified relative landmarks. The results show that the proposed method has characterization of shorter measure time and more flexible operation compared with the traditional size measurement method. Landmark recognition algorithm and body size measurement methods are efficient, and the rate of accuracy satisfies the demands of GB/T 23698—2009.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号