Generalized filter for removing video compression artifacts

A method comprising:

• obtaining a video signal to be encoded;
• encoding the video signal to obtain a two-dimensional array of reconstructed samples for each video frame of the video signal;
• filtering each target sample x(i,j) of the two-dimensional array of reconstructed samples based on values of samples in a neighboring region of the target sample x(i,j) to produce a two-dimensional array of modified reconstructed samples based on a first function that maps a value to an integer value in a range [0,2B−1], where B is a number of bits representing each sample of the two-dimensional array of modified reconstructed samples, and wherein the value is a sum of the target sample x(i,j) and a second function of a weighted sum over a neighboring region R around the target sample x(i,j) of a third function of a difference between the target sample x(i,j) and weighted values in the neighboring region R;
• and
• transmitting or storing an encoded video bit-stream based on the on the two-dimensional array of modified reconstructed samples for each video frame.

An apparatus comprising:

• a communication interface configured to receive an encoded video bit-stream to be decoded;
• and
• a processor coupled to the communication interface, wherein the processor is configured to:
• decode the encoded video bit-stream to obtain a two-dimensional array of reconstructed samples for each video frame in the encoded video bit-stream;
• and
• filter each target sample x(i,j) of the two-dimensional array of reconstructed samples based on values of samples in a neighboring region of the target sample x(i,j) to produce a two-dimensional array of modified reconstructed samples based on a first function that maps a value to an integer value in a range [0,2B−1], where B is a number of bits representing each sample of the two-dimensional array of modified reconstructed samples, and wherein the value is a sum of the target sample x(i,j) and a second function of a weighted sum over a neighboring region R around the target sample x(i,j) of a third function of a difference between the target sample x(i,j) and weighted values in the neighboring region R.