Reduced-Reference Quality Assessment of Point Clouds via Content-Oriented Saliency Projection

• DOI: 10.48550/arxiv.2301.07681
• Publication Date: 2023-01-01
• Authors: Wei Zhou; Guanghui Yue; Ruizeng Zhang; Yipeng Qin; Hantao Liu
• Keywords: Upsampling; Similarity (geometry)

Many dense 3D point clouds have been exploited to represent visual objects instead of traditional images or videos. To evaluate the perceptual quality various clouds, in this letter, we propose a novel and efficient Reduced-Reference metric for which is based on Content-oriented sAliency Projection (RR-CAP). Specifically, make first attempt simplify reference distorted into projected saliency maps with downsampling operation. Through process, tackle issue transmitting large-volume original user-ends assessment. Then, motivated by characteristics human system (HVS), objective scores are produced combining content-oriented similarity statistical correlation measurements. Finally, extensive experiments conducted SJTU-PCQA WPC databases. The experimental results demonstrate that our proposed algorithm outperforms existing reduced-reference no-reference metrics, significantly reduces performance gap between state-of-the-art full-reference assessment methods. In addition, show variation each technical component ablation tests.