The Research Of Youth Female Somatotype & 2D Non-contact Nomatometry System Based On Digital Pictures

Nowadays, the accuracy of anthropometric measurements is gradually important in pace with the increasing request of clothes'fitness. Because of its fast speed, convenient, accuracy and other merits, the non-contact measuring method has gradually taken the place of contact measuring method and become the major means of anthropometric measurement. The 3-dimensional non-contact somatometry system has the merits of direct, faster, precise and exhaustive. However, these 3-dimensional non-contact scanners are expensive, and inconveniently taken along. Besides, there are blind angles when measuring the body, which also bring about the resistance for measurement. Therefore, to develop a non-contact somatometry system that can be easily operated and lowly cost, and also be appropriate for the clothing industry has high practical value. The 2-dimensional non-contact somatometry system was established in this research. The merits of this system are lowly cost, be remote monitoring, easily operated, and can satisfy the meets of fast reaction and custom made on the internet to the modern clothing industry. The main researches are as follows:(1) The research object of this study is female body. The photogrammetry, hand dipping and 3D body scanner measurement are used to measure the body's sizes.(2) The shoulder angle, the body side angle, chest angle and hip angle are the characteristic angles for the somatotype. The female are classified into 64 sorts. The 10 main sorts'bodily form characteristics are analyzed.(3) The front and side faces of body's digital images are processed by the optimal threshold segmentation algorithm, filling, opening etc. Though those ways, the body edge can be extracted from the dark background pictures.(4) The computer graphics combined with the regular characteristics of bodies are used to determine the characteristic points on the body edge. The sizes can be extracted as follows: the body height, the neck thickness, the neck width, the shoulder width, the back width, the chest width, the bust width, the bust thickness, the BP points'width, the lowerberast thickness, the lowerbreast width, the waist thickness, the waist width, the abdomen thickness, the abdomen width, the hip thickness, the hip width, the thigh thickness, the thigh width, the knee thickness, the knee width, the calf thickness, the calf width and arm length.(5) The extracted values and manual measurements of shoulder width, back width, chest width and arm length are processed by regression analysis, and the arithmetic models are constructed.(6) The multiple linear regression analysis was done to construct the circumference arithmetic models in which the girth value measured by manual method is dependent variable and the thick and width values automatically extracted by the program are independent variables.(7) To establish the 2-dimensional non-contact somatometry system and the confirmation is done to this system.