摘要
非接触式三维人体自动测量是现代化人体测量技术的主要特征,是人体测量的发展趋势。采用kinect红外扫描仪进行人体着装测量,通过图像拼接技术获取人体3D图像,然后基于图像的人体特征区域和特征参数提取得到人体上衣定制的关键尺寸信息,包括肩宽、胸围、腰围、袖长、上领围和后中长。对由kinect红外扫描仪测量和手工测量的150组人体特征尺寸数据进行统计分析,建立两者的回归关系。结果表明,两者的测量误差小于±2 cm。根据GB 1335—1991《服装号型》国家标准,由kinect红外扫描仪提取的数据通过该回归方程转化后可用于定制服装或者推荐标准服装号型。
Non-contact three-dimensional human body automatic measurement is the main character of a modern measurementtechnology, and also is the development trend of body measurement. A kinect technique is used to scan human body in wearing state, and 3Dimages are obtained with image stitching technology, as well as human characteristics based on the image for customized clothing(includingshoulder width, bust width, waist width, sleeve length, neck around and centre of back). The 150 sets of characteristic size data by kinectIR scanner measurements and manual measurement are statistically analyzed, and the relationship between them is established by regressionmodal. Results show that the measuring error is less than 2cm according to the GB 1335—1991 standards of the garment, the measuring datafrom kinect IR scanner after converting by the regression equation can be used for customized clothing or recommending with standard models.
出处
《上海纺织科技》
北大核心
2016年第5期47-50,共4页
Shanghai Textile Science & Technology
关键词
人体测量
红外扫描仪
回归分析
标准
human body measurement
infrared scanner
regression analysis
standard