基于小波域特征提取的360度全景图像质量评价

摘要

With the development of Virtual Reality (VR)technology,whose representative 360-degree spherical images have a-roused an increasing number of attention.However,360-degree spherical images are easily distorted in the process of transmis-sion,compression,and store,so it is highly required to research how the popular coding technologies influence the quality of them.To this aim,the quality assessment of 360-degree spherical images is explored in this paper.A reduced-reference quality assessment method using feature information extracted in the wavelet domain,such as energy fluctuations,is designed for evalua-ting the quality of compressed 360 -degree spherical images.Compared with other image quality assessment algorithms,experi-ments verify that this method has higher performance.%随着虚拟现实(VR)技术的发展,其中代表性的360度全景图像越来越受到关注.本文针对360度全景图像在传输、压缩、储存过程中容易产生失真的特点,结合研究当下流行的编码技术对其质量影响的客观需求,研究了360 度全景图像的质量评价.设计了使用小波域提取能量波动等特征信息的半参考质量评价方法,用来对经过压缩的360 度全景图像进行质量评价.通过与其他图像质量评价算法结果的比较,验证了此方法具有较高的性能.