新型人体参数自动测量系统

发布时间：2018-05-04 21:43

本文选题：背景差分 + 分割　；参考：《中国图象图形学报》2017年02期

【摘要】：目的针对国内外已有的非接触式测量系统普遍存在价格昂贵、安装复杂、占用空间大等问题,提出一种基于消失点和图像比例法相结合的人体参数尺寸的自动测量方法(Hu FAMS-VP)。方法首先,利用背景差分法,将人体从图像中分割出来;再将分割后的人体图像进行轮廓提取,从而获得人体轮廓图;然后,根据边缘检测和人体各个部位与人体身高比例等方法,获取人体几个关键部位特征点;最后,通过消失点方法(vanishing point)与比例法的结合获取人体参数尺寸。结果根据Hu FAMS-VP对7组被测人员(年龄在25±3岁,身高范围在1.7±0.2 m)进行尺寸测量,并与被测人员真实尺寸信息进行比较。由测量结果可以得到7组被测人员的绝对误差控制在±0.05 m之内,平均测量速度在2 s以内。结论 Hu FAMS-VP在确保测量精度的前提下,具有操作方便、价格低廉等优势。根据测量结果,论证了本系统的准确性和鲁棒性。
[Abstract]:Objective to solve the problems of expensive price, complicated installation and large space occupation in non-contact measurement systems, an automatic measurement method of human parameter size based on the combination of vanishing point and image proportion method is proposed. Methods first of all, the human body is segmented from the image by using background difference method, and then the segmented human body image is extracted from the contour, and then the human body contour is obtained. According to the methods of edge detection and the ratio of human body height to human body height, several key human body feature points are obtained. Finally, the human body parameter size is obtained by the combination of vanishing point method and proportion method. Results according to Hu FAMS-VP, 7 groups of subjects (age 25 卤3 years, height range 1.7 卤0.2 m) were measured and compared with the real size information of the subjects. From the measurement results, we can get that the absolute error of the 7 groups is controlled within 卤0.05 m, and the average measuring speed is less than 2 seconds. Conclusion Hu FAMS-VP has the advantages of convenient operation and low price on the premise of ensuring measurement accuracy. According to the measurement results, the accuracy and robustness of the system are demonstrated.
【作者单位】：北京理工大学光电学院;
【基金】：国家自然科学基金项目(61572076) 中国博士后科学基金项目(2015M570940) 北京理工大学基础研究基金项目(20150442009)~~
【分类号】：TP391.41;TS941.17

【参考文献】