摘要
人体测量是通过测量人体各部位的尺寸来确定个体之间和群体之间在人体尺寸上的差别,用以研究人的形态特征。文中尝试一种兼顾精度、效率和成本的测量方式,即利用满足一定条件的人体照片,通过图像处理及后续搜索计算获得人体测量的数据。该文是在研究图像处理的部分理论基础上,重点分析了多种常用的边缘检测和轮廓提取算法,结合人体结构特点,设计了一个包括图像去噪、边缘检测、轮廓提取、尺寸顶点搜索和坐标计算的处理流程,达到人体尺寸信息自动提取的目的。结果极大地提高原型的健壮性和准确性,得到了理想的人体尺寸信息,说明该方法具有可行性。
Body measurement identifies the differences on measurement among an individual and population by measuring some body parts. So as it stands, attempt an measurement, giving attention to accuracy, cost and efficiency, by making use of image processing from photos meeting given conditions. Firstly research same digital image processing theories, analyze usual edge detection and contour extraction algorithms especially. With taking body structure into account, it designs a work flow including demising, edge detection, contour extraction, measurement vertex search and coordinates computation, which can realize extracting body measurernent automatically. And the experimental results improve the prototype, get the proper data of body measurement. So this method is feasible.
出处
《计算机技术与发展》
2010年第2期48-51,共4页
Computer Technology and Development
关键词
人体测量
边缘检测
轮廓提取
最佳阈值分割
body measurement
edge detection
contour extraction
optimum threshold segmentation
