Human body model establishing method, system and equipment and storage medium

Abstract

The invention discloses a method for establishing a human body model, and the method comprises the steps: obtaining a two-dimensional image of a target human body; substituting the two-dimensional human body contour image into a first neural network subjected to deep learning to carry out regression of joint points; acquiring information such as a joint diagram of the target human body; substituting into a second neural network subjected to deep learning to carry out regression of human body posture and body shape parameters; obtaining three-dimensional human body parameters including three-dimensional human body action posture parameters and three-dimensional human body shape parameters; constructing a three-dimensional standard human body model, namely a basic mannequin, in combination with the mathematical model; inputting the plurality of groups of obtained basis and skeleton parameters into a standard three-dimensional standard human body model for fitting; and obtaining a three-dimensional human body model with the same posture and body shape as the target human body. According to the three-dimensional human body model generation method, the self-owned standard human body model is established, the body bases suitable for the body shape characteristics of the Asian human body are selected, and the parameters corresponding to the 20 body bases and the 170 bones are used, so that the three-dimensional human body model which is closer to the body shape of the Asian human body than a Mapp human body SMPL model and is better in independent operation and controllability is generated.

Description

technical field [0001] The invention belongs to the field of three-dimensional human body modeling, and in particular relates to a human body modeling method, in particular to a method, system, device and storage medium for establishing a human body model based on machine learning to extract relevant information from photos. Background technique [0002] With the development of Internet technology, online shopping is becoming more and more popular. Compared with shopping in physical stores, online shopping has the advantages of a wide variety of products and convenient shopping. However, there are also some problems that are not easy to solve when buying goods online. The most important thing is that it is impossible to check the goods to be bought on the spot. Of all the product categories, this issue is most prominent for apparel products. Compared with shopping in physical stores where you can change clothes in real time to check the effect of clothing, online clothing ...

AI Technical Summary

Problems solved by technology

In this modeling method, the parameters that are judged by the model and finally output by the neural network are only the parameters of the relevant nodes, and then the mature body shape of the SMPL model is used to make detailed adjustments consistent with the target human body posture. Although the amount of calculation is reduced, due to There are few input parameters, and the adjustment can only be completed on the basis of the SMPL prediction model, so it is difficult to output a particularly ideal human body model that is highly consistent with the target human body posture

Images