System for video quality prediction and related aspects

A method of generating a measure of quality for a video signal representative of a plurality of frames, the video signal having:

• an original form;
• an encoded form in which the video signal has been encoded using a compression algorithm utilising a variable quantiser step size such that the encoded signal has a quantiser step size parameter associated therewith and utilising differential coding such that the encoded signal contains representations of the prediction residual of the signal;
• and a decoded form in which the encoded video signal has been at least in part reconverted to the original form, the method comprising:
• generating a first quality measure which is dependent on said quantiser step size parameter according to a predetermined relationship;
• generating a masking measure, the masking measure being dependant on the spatial complexity of at least part of the frames represented by the video signal in the decoded form according to a first predetermined relationship;
• and
• generating a combined measure, the combined measure being dependent upon both the first quality measure and the masking measure according to a predetermined relationship,
• wherein, the method is characterised by said masking measure being dependant on the spatial complexity of one or more active region(s) according to a predetermined relationship and on the spatial complexity of one or more background regions according to a predetermined relationship,
• wherein said one or more active regions are identified by:
• generating a first measure which is dependent on the prediction residual of the signal according to a first predetermined relationship, and
• identifying one or more regions for which the first measure meets a condition as one or more active regions;
• and,
• wherein said one or more background regions are identified by:
• generating a second measure which is dependent on the prediction residual of the signal according to a second predetermined relationship;
• and
• identifying one or more background regions of the picture for which the second measure meets a condition as one or more background regions.

Apparatus arranged to generate a measure of quality for a plurality of frames of a decoded form of a video signal, the video signal having:

• an original form;
• an encoded form in which the original form of the video signal has been encoded using a compression algorithm utilising a variable quantiser step size such that the encoded signal has a quantiser step size parameter associated therewith and utilising differential coding such that the encoded signal contains representations of the prediction residual of the signal;
• and said decoded form in which the encoded video signal has been at least in part reconverted to the original form, the apparatus comprising:
• means to generate a quality measure which is dependent on said quantiser step size parameter according to a predetermined relationship;
• means to determine one or more active regions in a picture of said video sequence for which the prediction residual exceeds a threshold condition;
• means to determine other types of regions within any remaining part of said picture;
• means to generate a masking measures dependent on the spatial complexity of each of said active regions and each determined type of region;
• and
• means to generate a combined quality measure by combining said quality measure and masking measure according to a predetermined relationship.