Surface fitting approach for reducing blocking artifacts in low bit-rate DCT decoded images

The loss of correlation in a signal in adjacent blocks due to the global nature of discrete cosine transform and the quantization of its coefficients are responsible for the blocking artifacts in low bit-rate block-based decompressed JPEG images. At very low bit-rates, these artifacts become visibly annoying, particularly in smooth areas or the image. Most de-blocking techniques to date are block-based and have only limited visual success because of the highly regular grid structure or the block boundaries. This paper is concerned mainly with the reduction of blocking artifacts in these smooth areas using a ternary polynomial surface to model the pixel intensity over a 2×2 block array called macro-block. Unlike current polynomial surface fitting techniques, including the JPEG technique, our novel technique attempts to compensate for the quantization error in the JPEG compressed DC coefficients. The technique uses a linear programming approach to optimize the fitting of the surface model to the quantized DC coefficients to within their corresponding JPEG quantization range. The proposed technique produces a much more pleasing image while having practically the same PSNR as the current de-blocking techniques using surface fitting approach