Research On Distortion Correction Algorithm For Virtual Reality Integrated Machine

With the advent of the 5G era,the continuous maturity and popularity of 5G technology will greatly promote the development of the virtual reality industry.Virtual reality devices are also receiving more and more attention,and various virtual reality products are gradually moving from the laboratory to ordinary users.For now,a good sense of virtual reality experience is still the goal pursued by related product manufacturers.A good experience doesn't only require the product to provide a strong sense of immersion,but also reduce the user's vertigo.This paper attempts to reduce the user's dizziness by studying software-level algorithm optimization for the user's vertigo caused by hardware design defects in the virtual reality integrated machine(VR integrated machine).The main causes of vertigo are distortion and dispersion.The main research point of this paper is image distortion in VR machine.In order to solve this problem,this paper has carried out the following research based on the existing research:(1)For the image distortion problem in VR integrated machine,this paper proposes a genetic algorithm based distortion correction optimization algorithm.In this paper,we use the division model proposed by Fitzgibbon.After studying and comparing the distortion parameters of the model in other lens equipment,we use the genetic algorithm to optimize the fitting distortion parameters,without limiting the image content,without preprocessing.Accurately and robustly eliminate image distortion in VR devices.At the same time,the distortion parameters in the model can be adaptively adjusted and can be applied to different scene distortion corrections.The experimental results show that the proposed method is more robust than the commonly used methods and the distortion correction effect is better.(2)Based on the real-time problem of distortion optimization algorithm,a modified MobileNet network model is constructed.The SE module of the Squeeze-and-Excitation network is used to embed the MobileNet network structure,and the VR scene is accurately identified and classified through the pretrained and improved lightweight network MobileNet,and then the scene is called according to different scenarios.Distortion parameters effectively correct the image to achieve the purpose of accurate distortion correction in real time in the virtual reality integrated machine.In this paper,the research on the image distortion correction algorithm in the virtual reality integrated machine can correct or even eliminate the image distortion existing in the device in real time,so that the user can feel the dizziness when using the device and improve the overall experience of the device.