A new Blocking Artifact Reduction Algorithm Using the Adaptive Filter Based on the Classification of the Block Boundary Area

摘要

In this paper, we proposed a new blocking artifact reduction algorithm using an adaptive filter based on the classification of the block boundary area. Generally, the block-based coding such as JPEG, MPEG introduce the blocking artifact and the ringing artifact to the image. The blocking artifact consists of the grid noise, the staircase noise, and the corner outlier. In the proposed method, the staircase noise and the corner outlier are reduced by lD lowpass filter. Then we classify the block boundaries into two classes based on the gradient of pixel intensity in the boundary region. For each class, adaptive filter is applied so that the grid noise is reduced at the block boundary regions. Next, for the block having the edge component, the ringing artifact is removed by applying the adaptive filter around the edge. Finally the high frequency components are added to only the block boundary in which the natural characteristic is lost by the adaptive filter. The performance of the proposed method is investigated by computer simulation in comparison with the traditional method. In the experimental results, we confirmed the better performance in the subjective and objective image quality.