文摘
The block-based DCT compression methods usually result in discontinuities called blocking artifacts at the boundaries of blocks due to the coarse quantization of the coefficients. In this paper, a new blocking artifacts reduction algorithm is proposed to reduce blocking artifacts and preserve edges adequately. We introduce the definition of homogeneous block and the measurement with an alterable threshold for judging whether a shifted block is homogeneous or inhomogeneous. For each row of the homogeneous blocks, an adaptive Sigmoid function is used to replace the step function to reduce blocking artifacts. For the inhomogeneous blocks, Sigma filter is used to smooth two adjacent blocks’ boundary. Simulation results evaluated by both visual and numerical quality metrics show that the proposed blocking artifacts reduction algorithm outperforms the original algorithms.