A Weighted Sobel Operator-Based No-Reference Blockiness Metric

A weighted Sobel operator-based no-reference perceptual blockiness metric (WSBM) which is used to measure blocking artifacts of DCT coding image, is presented in this paper. The proposed metric calculates the image luminance gradient matrix using Sobel operator. Based on the matrix, it quantifies the severity of blocking artifacts and the annoyance of large flatness in low bit-rate images. The new approach also takes into account the spatial human visual system (HVS) features, which include the luminance adaptation and texture masking. These processes are individually conducted block by block. Experimental results on various JPEG coded images demonstrate that the new metric is efficient in terms of computational complexity, and its performance is consistent with the subjective evaluation.