Motion compensation predicting encoding method and apparatus

Abstract

A motion compensation predicting coding apparatus includes a motion detector for comparing pixel data in an image signal with pixel data of the previous image signal to generate a motion vector. The motion vector is indicative of the displacement differences between the image signal and the previous image signal. The image signal may be organized into blocks of pixel data for a single field or for multiple fields when notion detection is performed on the image signal by the motion detector. A subtracter generates a prediction compensation error signal by subtracting the pixel data in the previous image signal from the pixel data of the image signal. A coder receives the prediction compensation error signal and generates an appropriate coded output. The apparatus may also include an adaptive blocking mechanism for receiving the prediction compensation error signal and organizing the data in the signal into blocks into one of several fashions as dictated by the motion vector.

We claim:  
     
       1. A motion compensation predicting coding apparatus for receiving odd and even fields of an input image signal, forming pixel data for each odd and even field into a block to obtain a current block image pattern, comparing the current block image pattern with a previous block image pattern to obtain a motion vector for the block, an obtaining a motion compliance prediction signal based upon the motion vector, said apparatus comprising: 
       a) a motion vector detector for generating a motion vector indicative of a displacement difference of the current block image pattern in one of two odd fields, two even fields, and an odd field and an even field, relative to the previous block image pattern;  
       b) a motion compensation prediction signal generator for generating the motion compensation prediction signal for each block of a field to be encoded; based upon said motion vector;  
       c) means for generating a prediction error signal for each block of a field to be encoded by subtracting said motion compensation prediction signal from the input image signal of said field to be encoded;  
       d) a coder for coding the prediction error signal, said coder including  
       i) a transformation means for performing an orthogonal transformation on the prediction error signal to produce a coefficient matrix that includes a plurality of coefficients in a single horizontal-vertical plane of the coefficient matrix;  
       ii) a scanning controller for determining a scanning order based on the motion vector, and for scanning said plurality of coefficients in the single horizontal-vertical plane of said coefficient matrix in accordance with the scanning order to read out said plurality of coefficients; and  
       iii) a quantizer for quantizing said plurality of coefficients rad out from said coefficient matrix; and  
       e) adaptive blocking means, coupled to the motion vector detector, for providing a composed error signal, said adaptive blocking means receiving the motion vector and responsively arranging therewith the composed error signal according to one of a field independent mode and a field composition mode to produce the composed error signal in which in the field composition mode, the composed error signal is arranged into alternating lines of data corresponding to the odd and even fields of the input image, and in which the in field independent mode, the composed error signal is arranged into a group having data corresponding to one field of the input image followed by a group having data corresponding to other field.  
     
     
       2. The motion compensation predicting coding apparatus of  claim 1 , further comprising: 
       a frame memory, coupled to the motion vector detector and to the means for generating a prediction error signal, for storing data corresponding to the previous block image pattern, wherein the frame memory includes producing means for producing a motion compensation prediction signal based on the motion vector and for providing the motion compensation prediction signal to the means for generating a prediction error signal.  
     
     
       3. The motion compensation predicting coding apparatus of  claim 1 , wherein the scanning order is determined to be one of a plurality of scanning orders, the plurality of scanning orders including: 
       a first order of scanning the plurality of coefficients in a horizontal direction of the coefficient matrix;  
       a second order of scanning the plurality of coefficients in a vertical direction of the coefficient matrix; and  
       a third order of scanning the plurality of coefficients in a diagonal direction of the coefficient matrix.  
     
     
       4. A motion compensation predicting coding apparatus for encoding pixel data of an image signal, wherein said image signal includes pixel data organized into an even field and an odd field for a portion of an image, said apparatus comprising: 
       a) a motion detector for comparing the pixel data of the image signal with pixel data of a previous image signal of a previous image, the motion detector including means for generating a motion vector indicative of displacement differences between the image signal and the previous image signal of the previous image;  
       b) a subtracter for generating a prediction compensation error signal by subtracting the pixel data of the previous image signal from the pixel data of the image signal;  
       c) an adaptive blocking mechanism for receiving the prediction compensation error signal and generating a composed signal, said adaptive blocking mechanism receiving the motion vector and responsively arranging therewith the composed signal according to one of a field independent mode and a field composition mode, in which the field composition mode, the error signal is arranged into alternating lines of data corresponding to the odd and even fields of the input image, and in which in the field independent mode, the error signal is arranged into a group having data corresponding to one field of the input image followed by a group having data corresponding to other field; and  
       d) a coder for receiving and encoding the composed signal from the adaptive blocking mechanism.  
     
     
       5. A motion compensation predicting decoding apparatus for receiving multiplexed coded data, the apparatus comprising: 
       
         a separator for separating the multiplexed coded data into a plurality of data blocks including a plurality of transformation factors, and a factor scanning order code defining the scanning order of the transform factors between alternative scanning orders, collectively defining a block prediction error signal, and a block structure code and motion vector associated with the block prediction error signal;  
       
       
         a decoder for decoding the prediction error from the plurality of transformation factors based on a scanning order defined by the factor scanning order code;  
       
       
         a blocking controller for determining a block type of each decoded block prediction error signal based on the block structure code;  
       
       
         a memory for storing an image signal, the image signal being read out therefrom shifted by the motion vector to form a motion compensated prediction signal; and  
       
         an adder for adding the decoded block - controlled block prediction error signals to the motion compensated prediction signal to produce a decoded image signal.   
     
     
       6. A motion compensation predicting decoding apparatus for receiving and decoding multiplexed coded data, the apparatus comprising: 
       
         a separator for separating the multiplexed coded data into a plurality of coded data blocks and associated encoding information, said encoding information including a factor scanning order defining the scanning order of transform factors within the coded data block between alternative scanning orders, a block structure code identifying whether each block is a field composed or a field independent block and a motion vector associated with the block prediction error within the encoded data blocks;  
       
       
         a decoder, responsive to said separator, for decoding each of the plurality of encoded data blocks based on the scanning order defined by the factor scanning order to produce predictively coded image information;  
       
       
         a blocking controller, receiving the output of the decoder, for determining a block type of each decoded block of predictively coded image information and assembling the decoded blocks in dependence on whether the blocks are field composed or field independent based on the block structure code;  
       
       
         a memory for storing an image signal; and  
       
       
         an adder, receiving the output of the blocking controller and the memory, for adding the predictively encoded image information to associated portions of the image signal stored in said memory. 
       
     
     
       7. The apparatus of  claim 6  wherein said encoded data blocks are orthagonal transformations. 
     
     
       8. A motion compensation predicting decoding apparatus for receiving and decoding multiplexed coded data, comprising: 
       
         a separator for separating the multiplexed coded data into a plurality of coded data blocks and associated encoding information, said encoding information including a factor scanning order defining the order of transform factors within the coded data block, a block structure code identifying whether each block is a field composed or a field independent block and a motion vector associated with the block prediction error within the encoded data blocks;  
       
       
         a decoder, responsive to said separator, for decoding each of the plurality of encoded data blocks based on the scanning order defined by the factor scanning order to produce predictively coded image information;  
       
       
         a blocking controller, receiving the output of the decoder, for determining a block type of each decoded block of predictively coded image information and assembling the decoded blocks in dependence on whether the blocks are field composed or field independent based on the block structure code;  
       
       
         a memory for storing an image signal; and  
       
       
         an adder, receiving the output of the blocking controller and the memory, for adding the predictively encoded image information to associated portions of the image signal stored in said memory;  
       
       
         wherein said block structure code is defined by the motion vector. 
       
     
     
       9. The apparatus of  claim 8  wherein said factor scanning order is defined by the motion vector.