The invention discloses a mobile terminal surface light field rendering method based on deep learning

Abstract

The invention discloses a mobile terminal surface light field rendering method based on deep learning. The mobile terminal surface light field rendering method comprises the steps that S1, a camera isused for shooting sampling pictures of a rendered target object at all angles towards the target object; S2, synthesizing a 360-degree environment map of the current environment of the target objectby using the picture; Step S3, aligning the sampled photo to a space coordinate system consistent with the model file of the rendered target object; S4, collecting training data of the deep neural network, and building and training the deep neural network; And S5, rendering the representation of the rendered target object at any viewing angle in real time, and realizing the switching of 360-degreeenvironment maps to obtain the effects of different rendering results. According to the method, an advanced deep learning framework is applied, an integral in a rendering equation is fitted, the calculation amount in the rendering process is reduced, during rendering, environment chartlets can be switched at will, and a relay effect is achieved.

Description

technical field [0001] The invention relates to the field of object surface rendering, in particular to a deep learning-based light field rendering method for mobile terminal surfaces. Background technique [0002] The research related to recreating the appearance of objects in the real world has always been an active research direction in the field of computer graphics and computer vision, because this technology has great practical application scenarios. In addition to virtual reality and augmented reality, the industry Clothing material design can also use this technology to simulate and emulate virtual materials, and the e-commerce field can also use this technology to truly display products to users. In recent years, some mobile devices such as smartphones and some head-mounted devices have developed rapidly. Most of the devices are equipped with cameras and GPUs (graphics processing units). Such devices make object rendering on the mobile side more reliable and feasibl...

AI Technical Summary

Problems solved by technology

[0006] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is to overcome the problems in the prior art such as the large amount of calculation of the object surface rendering method, and provide a new method for reappearing the appearance of an object by means of a deep network

Images