Objective Quality Assessment Metrics for Light Field Image Based on Textural Features

Light Field (LF) imaging is a plenoptic data collection method enabling wide variety of image post-processing such as 3D extraction, viewpoint change and digital refocusing. Moreover, LF provides the capability to capture rich information about scene, e.g., texture, geometric information, etc. Therefore, quality assessment model for images needed poses significant challenges. Many Image Quality Assessment (LF-IQA) metrics have been recently presented based on unique characteristics images. The state-of-the-art objective taken into account content human visual system SSIM IW-SSIM. However, most these are designed video with natural content. Additionally, other models (e.g., depth angle information) trade high performance computational complexity, along them possessing difficulties implementation applications due immense requirements Hence, this paper presents novel content-adaptive LF-IQA metric improve conventional that also low in complexity. experimental results clearly show improved compared IQA metrics, we identify well-suited assessment. In addition, present comprehensive content-based feature analysis determine appropriate influences perception among widely used metrics. Finally, dataset selected from EPFL dataset, allowing study light field by qualitative factors (wide narrow), focus (background or foreground) complexity (simple complex).