Three-dimensional human body modeling method based on OpenGL and deep learning

Abstract

The invention discloses a three-dimensional human body modeling method based on OpenGL and deep learning. The method comprises the following steps: extracting a human body image in a two-dimensional image through a deep learning model Mask R-CNN, and reconstructing a personalized three-dimensional human body model through the combination of OpenGL and a standard 3D human body model. The method comprises the following steps: firstly, segmenting a human body image in a two-dimensional image by adopting a Mask R-CNN deep learning model; and then extracting main features of the human body contourobtained after segmentation, finally mapping the features of the human body contour image to a three-dimensional standard human body model established by 3d-max by using OpenGL, and quickly constructing a three-dimensional human body model in the OpenGL. The method is fast in image processing and high in model generation efficiency.

Description

technical field [0001] The invention relates to a three-dimensional human body model established by using a deep learning model and OpenGL, and belongs to the technical field of three-dimensional human body modeling. Background technique [0002] The generation of 3D human body models refers to the creation of digital geometric models of human body objects in virtual scenes. Generating a high-fidelity 3D human body model with individual shape characteristics is a classic problem that has been studied since the birth of computer graphics, and it is still a hot issue in academia and industry today. There are two reasons for this fact: On the one hand, the 3D human body model has a high complexity. On the other hand, users can easily perceive the degree of distortion of the generative model. [0003] Traditional 3D model generation technology is difficult to provide practical human body modeling methods to mass users. The current 3D model generation technology can be mainly ...

AI Technical Summary

Problems solved by technology

The first method is limited by the high price of 3D laser scanners, and is currently generally used by laboratories or companies

The 3D editing software in the second method is generally oriented to professional 3D model design. Making a 3D model with individual shape features requires professional artists and consumes a lot of manpower, material resources and financial resources.

The shortcomings of the above two typical technologies severely limit the generation of 3D human body models, and ordinary users cannot easily obtain their own 3D models

Images