Apparatus and method for optimized compression of interlaced motion images

Abstract

An interlaced image processing module and corresponding method facilitate improved processing of interlaced motion images. In one embodiment, the interlaced image processing module receives image data frames having interlaced first and second fields and produces a reference field and error field. The reference field corresponds to the still image content of the interlaced frame, whereas the error field corresponds to the motion content of the interlaced frame, particularly the motion between fields. Motion between fields is thus represented in the error field, without redundant representation of the still image content provided by the first field. Where there is little motion between fields, the error terms will be small so the predictor preserves the coding efficiency provided by any auto-correlation in the image. Further, the interlaced image processing method does not rely upon pixel group classification, and thus avoids classification errors, and the loss of coding efficiency from still image content in motion classified blocks. Finally, problems presented by relative motion between fields are avoided, as are local artifacts. Another embodiment transforms the interlaced fields into frame data having a high frequency field and a low frequency field.

I claim:  
     
         1 . A method for compressing an image signal, the method comprising: 
 receiving image data corresponding to a plurality of neighboring fields of the image signal;    obtaining, using a wavelet transform operating on the image data corresponding to the plurality of neighboring fields, a frequency-partitioned data set comprising a plurality of frequency bands.    
     
     
         2 . A method according to  claim 1 , wherein the wavelet transform is a three-dimensional wavelet transform.  
     
     
         3 . A method according to  claim 1 , wherein the method comprises processing video data that is a mixture of interlaced and progressive format video.  
     
     
         4 . A method for reverse processing of a compressed image signal, the method comprising: 
 receiving image data corresponding to a frequency-partitioned data set comprising a plurality of frequency bands;    obtaining, using a reverse wavelet transform operating on the frequency-partitioned data set, a plurality of neighboring fields of the image signal.    
     
     
         5 . A method according to  claim 4 , wherein the reverse wavelet transform is a three-dimensional wavelet transform.  
     
     
         6 . A method according to  claim 4 , wherein the method comprises reverse processing of a compressed image signal that is a mixture of interlaced and progressive format video.  
     
     
         7 . A method for reverse processing of a compressed image signal, the method comprising: 
 receiving, for each frame of the image signal, image data corresponding to a low frequency field and a high frequency field;    using a reverse wavelet transform to obtain a first field for each frame, based on the frame's low frequency field and high frequency field;    using a reverse wavelet transform to obtain a second field for each frame, based on the frame's low frequency field and high frequency field; and    producing, for each frame, frame data comprising the frame's first field and second field, the first field and the second field being interlaced.    
     
     
         8 . A method according to  claim 7 , in which the reverse wavelet transform is a two-dimensional wavelet transform.  
     
     
         9 . A method according to  claim 8 , in which the reverse wavelet transform is a nonseparable, near orthogonal wavelet transform.  
     
     
         10 . A method according to  claim 9 , in which the reverse wavelet transform is implemented as an odd-sized linear filter.  
     
     
         11 . A method according to  claim 8 , in which the reverse wavelet transform is a biorthogonal wavelet transform.  
     
     
         12 . A method according to  claim 11 , in which the reverse wavelet transform is a 3-by-3 linear filter.  
     
     
         13 . A method according to  claim 7 , in which the reverse wavelet transform is a spatial-only wavelet transform.  
     
     
         14 . A method according to  claim 13 , in which the reverse wavelet transform is a nonseparable, near orthogonal wavelet transform.  
     
     
         15 . A method according to  claim 13 , in which the reverse wavelet transform is a biorthogonal wavelet transform.  
     
     
         16 . A method according to  claim 7 , in which the method comprises processing video data that is a mixture of interlaced and progressive format video.  
     
     
         17 . An apparatus for reverse processing of a compressed image signal, the apparatus comprising: 
 an input for receiving, for each frame of the image signal, image data corresponding to a low frequency field and a high frequency field;    a reverse wavelet transform module, in communication with the input, for obtaining, using a reverse wavelet transform, a first field for each frame based on the frame's low frequency field and high frequency field, and for obtaining, using a reverse wavelet transform, a second field for each frame based on the frame's low frequency field and high frequency field; and    a signal merging module, in communication with the reverse wavelet transform module, for producing, for each frame, frame data comprising the frame's first field and second field, the first field and the second field being interlaced.    
     
     
         18 . An apparatus according to  claim 17 , in which the reverse wavelet transform module implements a two-dimensional wavelet transform.  
     
     
         19 . An apparatus according to  claim 18 , in which the reverse wavelet transform module implements a nonseparable, near orthogonal wavelet transform.  
     
     
         20 . An apparatus according to  claim 19 , in which the reverse wavelet transform module implements an odd-sized linear filter.  
     
     
         21 . An apparatus according to  claim 18 , in which the reverse wavelet transform module implements a biorthogonal wavelet transform.  
     
     
         22 . An apparatus according to  claim 21 , in which the reverse wavelet transform module implements a 3-by-3 linear filter.  
     
     
         23 . An apparatus according to  claim 17 , in which the reverse wavelet transform module implements a spatial-only wavelet transform.  
     
     
         24 . An apparatus according to  claim 23 , in which the reverse wavelet transform module implements a nonseparable, near orthogonal wavelet transform.  
     
     
         25 . An apparatus according to  claim 23 , in which the reverse wavelet transform implements a biorthogonal wavelet transform.  
     
     
         26 . An apparatus for reverse processing of a compressed image signal, the apparatus comprising: 
 means for receiving, for each frame of the image signal, image data corresponding to a low frequency field and a high frequency field;    means for obtaining, using a reverse wavelet transform, a first field for each frame based on the frame's low frequency field and high frequency field;    means for obtaining, using a reverse wavelet transform, a second field for each frame based on the frame's low frequency field and high frequency field; and    means for producing, for each frame, frame data comprising the frame's first field and second field, the first field and the second field being interlaced.    
     
     
         27 . An apparatus according to  claim 26 , in which the means for obtaining a first field and the means for obtaining a second field comprise means for performing a two-dimensional wavelet transform.  
     
     
         28 . An apparatus according to  claim 27 , in which the two-dimensional wavelet transform means comprises means for performing a nonseparable, near orthogonal wavelet transform.  
     
     
         29 . An apparatus according to  claim 28 , in which the nonseparable wavelet transform means comprises an odd-sized linear filter.  
     
     
         30 . An apparatus according to  claim 27 , in which the two-dimensional wavelet transform means comprises means for performing a biorthogonal wavelet transform.  
     
     
         31 . An apparatus according to  claim 30 , in which the biorthogonal wavelet transform means comprises a 3-by-3 linear filter.  
     
     
         32 . An apparatus according to  claim 26 , in which the means for obtaining a first field and the means for obtaining a second field comprise means for performing a spatial-only wavelet transform.  
     
     
         33 . An apparatus according to  claim 32 , in which the spatial-only wavelet transform means comprises means for performing a nonseparable, near orthogonal wavelet transform.  
     
     
         34 . An apparatus according to  claim 32 , in which the spatial-only wavelet transform means comprises means for performing a biorthogonal wavelet transform.