Compensating quantization errors of a decoded video signal by using an adaptive filter

摘要

The image quality of a decoded video signal is improved by using an adaptive filter 150. Specifically, a transmitted encoded video signal is converted to a set of discrete cosine transform coefficients corresponding to current frame data through the use of variable length decoding 130 and inverse quantization 140 techniques. High frequency components of the set of discrete cosine transform are adaptively filtered 150 based on a spatial activity of current frame data determined by utilizing previous frame data decoded prior to deriving the current frame data. Then, a set of filtered discrete cosine transform coefficients is transformed 160 into a set of inverse discrete cosine transform data. The set of inverse discrete cosine transform data and the previous frame data are used to produce a current frame data through the use of motion compensation 170-190. The current frame data serves as new previous frame data and the decoded video signal at the same time.