Blind Panoramic Image Quality Assessment via the Asymmetric Mechanism of Human Brain
Yumeng Xia, Yongfang Wang, Peng Ye · 2019
Most existing objective quality evaluation metrics for panoramic images are typically derived from peak signal to noise ratio (PSNR) or structural similarity (SSIM). However, they need pristine panoramic images and are not highly consistent with human perception. To address the problem, we propose a novel blind panoramic image quality assessment (PIQA) method to predict the visual quality of panoramic images based on asymmetric mechanism of human brain. In the proposed method, the high-frequency feature is extracted by panoramic-weighted local binary pattern and relative total variation is used to clip the high-frequency information in panoramic images, then panoramic-weighted statistic feature can represent the lowfrequency feature. Finally, support vector regression (SVR) is adopted to build a quality predictor from feature space to quality score space. The experimental results on the public subjective dataset demonstrate the superiority of our proposed metric compared with state-of-the-art objective PIQA methods.