Video companding method and system

A method in a signal processing system for processing of video data comprising performing a decode process, for i running from 1 to m where m is an integer larger than 1 and corresponds to a number of portions of a video frame by the steps of:

• recovering a signal F.sub.i from a medium, signal F.sub.i corresponding to a signal E.sub.i as altered by noise and/or distortion, if any, which said signal E.sub.i was previously applied to the medium after performing a forward encode process and corresponds to a portion P.sub.i of a video frame, the decoder process being the inverse of the forward encode process;
• subtracting a local average G.sub.i of said signal F.sub.i from said signal F.sub.i to generate a signal H.sub.i ;
• adding an offset R.sub.i to said signal H.sub.i to generate a signal J.sub.i ;
• applying a non-linear reconstruction process to said signal J.sub.i to generate a signal K.sub.i, said non-linear reconstruction process being monotonic and having a transfer curve with a relatively lower slope around the origin and a relatively higher slope further away from the origin.
• adding said local average G.sub.i to said signal K.sub.i, the sum of said local average G.sub.i and said signal K.sub.i being signal L.sub.i ;
• and
• supplying said signal L.sub.i to a video monitor to display a video frame with relatively low contrast parts of the video frame having an increased signal-to-noise ratio, the increased signal-to-noise ratio for a particular low contrast part being increased relative to a signal-to-noise ratio for that low contrast part in the absence of the forward encode process, and relatively high contrast parts of the video frame having a decreased signal-to-noise ratio, the decreased signal-to-noise ratio for a particular high contrast part being decreased relative to a signal-to-noise ratio for that high contrast part in the absence of the forward encode process.

A signal processing system for processing of video data comprising means for performing a forward encode process including:

• means for calculating a local average A.sub.i of a portion V.sub.i of a video input V, which portion corresponds to a portion P.sub.i of a video frame, for i running from 1 to m where m is an integer larger than 1 and is the number of portions P.sub.i of the video frame for which the local average A.sub.i is calculated;
• means, connected to said means for calculating, for subtracting said local average A.sub.i from V.sub.i to generate a signal B.sub.i ;
• means, connected to said means for calculating, for applying a non-linear forward process to generate a signal C.sub.i which emphasizes any values of B.sub.i corresponding to any values of V.sub.i having relatively small differences from the local average A.sub.i and which de-emphasizes any values of B.sub.i corresponding to any values of V.sub.i having relatively large differences from the local average A.sub.i ;
• and
• means, connected to said means for applying, for using said signal C.sub.i to provide and apply an encoded output E.sub.i to a medium such that said encoded output E.sub.i is recoverable from said medium to reconstruct the video frame with relatively low contrast parts of the video frame having an increased signal-to-noise ratio, the increased signal-to-noise ratio for a particular low contrast part being increased relative to a signal-to-noise ratio for that low contrast part in the absence of the forward encode process, and relatively high contrast parts of the video frame having a decreased signal-to-noise ratio, the decreased signal-to-noise ratio for a particular high contrast part being decreased relative to a signal-to-noise ratio for that high contrast part in the absence of the forward encode process.

A signal processing system for processing of video data comprising means for performing a decode process including:

• means for recovering a signal F.sub.i from a medium, from i running from 1 to m where m corresponds to a number of portions of a video frame said signal F.sub.i corresponding to a signal E.sub.i as altered by noise and/or distortion, which said signal E.sub.i was previously applied to the medium after performing a forward encode process and corresponds to a portion P.sub.i of a video frame, the decode process being the inverse of the forward encode process;
• means for subtracting a local average G.sub.i of said signal F.sub.i from said signal F.sub.i to general a signal H.sub.i ;
• means for adding an offset R.sub.i to said signal H.sub.i to generate a signal J.sub.i ;
• means, connected to said means for recovering, for applying a non-linear reconstruction process to said signal J.sub.i to generate a signal K.sub.i, said non-linear reconstruction process being monotonic and having a transfer curve with a relatively lower slope around the origin and a relatively higher slope further away from the origin;
• means, connected to said means for applying, for adding said local average G.sub.i to said signal K.sub.i, the sum of said local average G.sub.i and said signal K.sub.i being signal L.sub.i ;
• and
• means, connected to said means for adding, for supplying said signal L.sub.i to a video monitor to display a video frame with relatively low contrast parts of the video frame having an increased signal-to-noise ratio, the increased signal-to-noise ratio for a particular low contrast part being increased relative to a signal-to-noise ratio for that low contrast part in the absence of the forward encode process, and relatively high contrast parts of the video frame having a decreased signal-to-noise ratio, the decreased signal-to-noise ratio for a particular high contrast part being decreased relative to a signal-to-noise ratio for that high contrast part in the absence of the forward encode process.