Study On The Extraction Method Of Online Parameter Of Type Human Body Based On Photos

Now with the network shopping popular, many people choose to buy their needed the clothes on the net, the network buys itself has many problems, such as people on the network can not be like the clothes in the store as personal style is appropriate and fitting to feel the size fits, user supplied its measurement data because the method is not necessarily correct body not very reliable, which bring many buyers receive clothes found inconsistent results and expected to return replacement. In this paper, in order to meet a lot of online clothing store in order to provide more accurate size clothes to the customer demand to provide a method to extract the shape parameters of the human body based on photos online. Different body posture photographs and background image online this method using HTML5technology through the front-end extraction, after the introduction of difference, two value of the extracted contour, then computing the measure of image feature information interaction between human relationship and pictures of different postures of the human body based on the pixel position corresponding to the clothes, calculate the size needed the size of the value, then according to the page the customers own given height value and calculate the scale, and then calculate the various parts of the actual size. The platform server background part to construct Node.js and MongoDB use; part of the calculation using the c++implementation and the OpenCV function library for auxiliary treatment, dynamic link library and the web through the c#invokes the c++compiler generated is investigated using Web Socket communication, to realize the calculation module.