Digital video format converter and method therefor
View Patent ↗A digital video format converter and method therefor includes a first and second converter. The first converter receives a digital video signal and converts, in the frequency domain, inter-coded pictures in the digital video signal into intra-coded pictures. The first converter performs motion compensation in the frequency domain on the inter-coded pictures in the digital video signal to convert the inter-coded pictures into intra-coded pictures. The second converter down converts output of the first converter. Further signal processing then completes placing the digital video signal output by the second converter in a format acceptable to standard lower definition digital decoders.
1. An apparatus for down-converting a digital video signal, comprising:
receiving means for receiving an encoded digital video signal, said encoded digital video signal having been previously generated by converting spatial video data into the DCT domain, quantizing DCT domain coefficients, and variable length encoding the quantized DCT domain coefficients;
a variable length decoder for generating variable length decoded DCT data from said received encoded digital video signal;
a dequantizer for dequantizing the variable length decoded DCT coefficients generated by said variable length decoder to generate at least first and second DCT blocks, said first and second DCT blocks each having first dimensions;
converting means for receiving said at least first and second DCT blocks, and for converting said first and second DCT blocks into a single DCT block having dimensions equal to said first dimensions;
a quantizer for re-quantizing the output of said converting means, the quantization scale used by said quantizer for re-quantizing being different than the quantization scale used to generate said encoded digital video signal; and
a variable length coder variable length coding output of said quantizer.
2. The apparatus of claim 1 , wherein said quantizer re-quantizes said output of said converting means such that distortion is eliminated.
3. A method for down-converting a digital video signal, comprising:
(a) receiving an encoded digital video signal, said encoded digital video signal having been previously generated by converting spatial video data into the DCT domain, quantizing DCT domain coefficients, and variable length encoding the quantized DCT domain coefficients;
(b) variable length decoding said encoded digital video signal to generate variable length decoded DCT data;
(c) dequantizing said variable length decoded DCT data to generate at least first and second DCT blocks, said first and second DCT blocks each having first dimensions;
(d) converting said first and second DCT blocks into a single DCT block having dimensions equal to said first dimensions;
(e) re-quantizing the output of said step (d), the quantization scale used by said quantizer for re-quantizing being different than the quantization scale used to generate said encoded digital video signal; and
(f) variable length coding the output of said step (e).
4. The method of claim 3 wherein said step (e) re-quantizes said output of step (d) using a quantization scale which does not cause distortion in the re-quantized data.