VR低质视频图像局部质量自动增强方法仿真

Simulation of Local Quality Automatic Enhancement Method for VR Low-Quality Video Images

由于(Virtual Reality, VR)视频是通过压缩编码等方式进行传输和存储的,在传输过程中其帧间的信息会出现丢失或失真的现象,导致视频质量降低。为了优化VR视频的视觉体验,提出VR低质视频图像质量局部自动增强方法。基于VR视频内部的全景图像拼接原理,引入VR视频缩放因子对球形显示空间内的图像展开客观质量评价。当判定全景图像质量较低时,采取局部质量增强处理。通过投影变换方法构建全景图像畸变模型,利用采样点公式计算出用于畸变矫正的相关参数,进而达到全景的几何矫正目的。采用自适应Gamma变换与对CLAHE算法对原始光照分量的自适应调整,以及调整后光照分量的融合,达到全景图像局部增强的目的。实验结果表明,所提方法能够在最大化保留原始图像细节的前提下,实现低质量VR视频图像的增强。

Because virtual reality(VR)video is transmitted and stored through compression coding and other methods,the information between its frames will be lost or distorted during transmission,leading to the degradation of video quality.In order to optimize the visual experience of VR video,a local automatic enhancement method of VR low-quality video image quality is proposed.Based on the principle of panoramic image mosaic in VR video,the VR video scaling factor is introduced to objectively evaluate the image quality in spherical display space.When the panoramic image quality is judged to be low,local quality enhancement processing is adopted.The distortion model of the panoramic image is constructed by the projection transformation method,and the relevant parameters for distortion correction are calculated by the formula of sampling points,so as to achieve the purpose of geometric correction of the panoramic image.Adaptive gamma transform and CLAHE algorithm are used to adjust the original lighting component adaptively,and the adjusted lighting component is fused to achieve the purpose of local enhancement of the panoramic image.Experimental results show that the proposed method can enhance low-quality VR video images while maximizing the details of the original image.