FS-BAND: A Frequency-Sensitive Banding Detector
This addresses a specific quality annoyance in image processing for users affected by compression artifacts, representing an incremental improvement in detection methods.
The paper tackled the problem of detecting banding artifacts in images, which degrade user experience, by proposing a no-reference detector that analyzes frequency aspects and generates pixel-wise maps with quality scores, achieving higher accuracy than state-of-the-art methods in classification tasks.
Banding artifact, as known as staircase-like contour, is a common quality annoyance that happens in compression, transmission, etc. scenarios, which largely affects the user's quality of experience (QoE). The banding distortion typically appears as relatively small pixel-wise variations in smooth backgrounds, which is difficult to analyze in the spatial domain but easily reflected in the frequency domain. In this paper, we thereby study the banding artifact from the frequency aspect and propose a no-reference banding detection model to capture and evaluate banding artifacts, called the Frequency-Sensitive BANding Detector (FS-BAND). The proposed detector is able to generate a pixel-wise banding map with a perception correlated quality score. Experimental results show that the proposed FS-BAND method outperforms state-of-the-art image quality assessment (IQA) approaches with higher accuracy in banding classification task.