Quickhull凸包算法在人体围度确定中的应用及改进
2021-08-09方琦孙光武陈郁
方琦 孙光武 陈郁



摘 要:面向人体三维轮廓测量的凸包算法是确定服装截面围度尺寸的有效工具。为了减少耗时提高计算效率,引入了Quickhull凸包算法,根据人体具有20的凹凸对称性特征,在构建初始凸包前进行了对已知凹点直接删除的改进,并通过对女性最大胸围截面为例进行了说明,与Graham凸包算法精度和效率的比较,结果表明:采用Quickhull算法可获得与Graham算法相同的尺寸精度,但耗时明显减少,而通过对人体已知凹点的直接删除则可进一步减少计算的耗时。
关键词:人体围度;凸包算法;Quickhull算法
中图分类号: TS941.17
文献标志码:A
文章编号:1009-265X(2021)04-0081-05
Abstract: The convex hull algorithm for measuring the three-dimensional contour of human body is an effective tool to determine the size of the clothing section. In order to reduce time consumption and improve computational efficiency, this paper introduced the Quickhull convex hull algorithm. Before constructing the initial convex hull, the method of directly deleting the known concave points was improved according to the concave-convex symmetry of the human body. The method is illustrated by an example of the maximum female chest section. The result of comparison with the accuracy and efficiency of Graham convex hull algorithm showed that, Quickhull algorithm could be used to obtain the same dimensional accuracy as the Graham algorithm, while the time consumption of the Quickhull algorithm decreased significantly. The time consumption could be further reduced by directly deleting the known concave points of human body.
Key words: body girth; convex hull algorithm; Quickhull algorithm
準确的人体轮廓尺寸是服装设计的基础,对于服装的结构和号型确定,款式和穿着舒适度都极为重要[1-3]。这些如围度、高度和相应角度等必要的数据,传统上主要采用手工测量的方法获取[4]。随着科学技术的进步,三维扫描的测量方法得到发展并获得越来越多的应用[5-6],与手工测量方法相比,这种非接触式的测量方法具有数据更多、尺寸更准和简便快捷的优势[7]。
非接触式的测量方法通过高精度扫描仪获得的人体点云数据还需要进一步采用凸包算法对围度尺寸加以确定。凸包算法是一种广泛用于图形处理方法,常用的有二等分法[8]、曲线定向法[9]、Graham算法[10]和Quickhull算法[11]等。目前服装业中应用的主要是Graham算法,这种方法在人体扫描获得的各截面点云数据基础上,通过对轮廓点单向递推和回溯的计算,逐步扣除凹点,最终获得可用于服装设计的各人体截面围度凸包[12-13]。……
