Method and apparatus for processing an image in an image compression/decompression system that uses hierachical coding

An apparatus for use in processing data representative of an image, the apparatus comprising:

• means for receiving data indicative of one or more reconstructed versions of the image including data representative of a causal context of an object pixel of the image;
• means for generating a plurality of predictions for the object pixel and a plurality of predictions for at least one pixel of the causal context;
• means for determining at least one measure of correlation between the plurality of predictions for the at least one pixel of the causal context and the data representative of the causal context;
• and
• means for determining a prediction for the object pixel computed as a weighted sum of one or more of the plurality of predictions for the object pixel, in accordance with the at least one measure of correlation and a weighting policy.

A method for use in processing data representative of an image, the method comprising the steps of:

• receiving data representative of a prediction of an object pixel;
• receiving data representative of a plurality of reconstructed versions of an image including data representative of two or more causal contexts of the image from two or more resolutions of the image;
• determining a context for the two or more causal contexts;
• determining a substantially mean error for the context;
• and
• determining a prediction for the object pixel in accordance with the data representative of the prediction for the object pixel and the substantially mean error for the context.

An apparatus for use in processing data representative of an image, the apparatus comprising:

• means for receiving data representative of a prediction of an object pixel;
• means for receiving data representative of a plurality of reconstructed versions of an image including data representative of two or more causal contexts of the image from two or more resolutions of the image;
• means for determining a context for the two or more causal contexts;
• means for determining a substantially mean error for the context;
• and
• means for determining a prediction for the object pixel in accordance with the data representative of the prediction for the object pixel and the substantially mean error for the context.

A method for use in processing data representative of an image, the method comprising the steps of:

• receiving data representative of one or more prediction differences versions of the image including data representative of a causal context prediction difference of an object predictions difference of the image;
• generating a plurality of predictions for the object prediction difference and a plurality of predictions for at least one prediction difference of the causal context;
• determining at least one measure of correlation between the plurality of predictions for the at least one prediction difference and the data representative of the causal context predictions difference;
• and
• determining a prediction for the object prediction difference computed as a weighted sum of one or more of the plurality of predictions for the object prediction difference, in accordance with the at least one measure of correlation and a weighting policy.

An apparatus for use in processing data representative of an image, the apparatus comprising:

• means for receiving data representative of one or more prediction differences versions of the image including data representative of a causal context prediction difference of an object prediction difference of the image;
• means for generating a plurality of predictions for the object prediction difference and a plurality of predictions for at least one prediction difference of the causal context;
• means for determining at least one measure of correlation between the plurality of predictions for the at least one prediction difference and the data representative of the causal context prediction difference;
• and
• means for determining a prediction for the object prediction difference computed as a weighted sum of one or more of the plurality of predictions for the object prediction difference, in accordance with the at least one measure of correlation and a weighting policy.

A method for use in processing data representative of an image, the method comprising the steps of:

• receiving data representative of a prediction of an object prediction difference;
• receiving data representative of a plurality of reconstructed versions of an image including data representative of two or more causal contexts of the image from two or more resolutions of the image;
• determining a context for the two or more causal contexts;
• determining a substantially mean error for the context;
• and
• determining a prediction for the object prediction difference in accordance with the data representative of the prediction for the object prediction difference and the substantially mean error for the context.

An apparatus for use in processing data representative of an image, the apparatus comprising:

• means for receiving data representative of a prediction of an object predictions difference;
• means for receiving data representative of a plurality of reconstructed versions of an image including data representative of two or more causal contexts of the image from two or more resolutions of the image;
• means for determining a context for the two or more causal contexts;
• means for determining a substantially mean error for the context;
• and
• means for determining a prediction for the object prediction difference in accordance with the data representative of the prediction for the object prediction difference and the substantially mean error for the context.