Fine granularity scalability encoding/decoding apparatus and method
A fine granularity scalability encoding/decoding apparatus includes a first quantizer for performing a DCT on a motion-compensated image and quantizing a resulting value, a second quantizer for re-quantizing the value obtained by the first quantizer, an inverse-quantizer for re-quantizing a value re-quantized by the second quantizer, and a number of subtracters. A first subtracter obtains a difference between a value of N times the re-quantized value and the inverse-quantized value. A second subtracter obtains a difference between the value quantized by the first quantizer and the value quantized by the inverse-quantizer. And, a third subtracter subtracts the output value of the first subtracter from the output value of the second subtracter. Coding efficiency is increased and a bit-rate reduction of an entire coding stream is achieved by making an enhancement layer signal and a base layer signal have the same code.
1. A fine granularity scalability encoding apparatus, comprising:
a first quantizer which quantizes a resulting DCT value from a motion-compensated image;
a second quantizer which re-quantizes the value obtained by the first quantizer;
an inverse-quantizer which inverse quantizes a value re-quantized by the second quantizer;
a first subtracter which obtains a difference between a value of N times the re-quantized value and the inverse-quantized value;
a second subtracter which obtains a difference between the value quantized by the first quantizer and the value quantized by the inverse-quantizer; and
a third subtracter which subtracts an output value of the first subtracter from an output value of the second subtracter.
2. The apparatus of claim 1 , further comprising:
a bit plane shifting unit which performs bit plane shifting if a block with a good picture quality is required.
3. The apparatus of claim 1 , further comprising:
a maximum value calculating unit which searches a maximum value from output values of the third subtracter; and
a bit plane variable length coding unit which performs variable length coding on the obtained maximum value by bit planes.
4. A fine granularity scalability encoding apparatus, comprising:
a first quantizer which quantizes a resulting DCT value from a motion-compensated image;
a second quantizer which re-quantizes the value obtained by the first quantizer;
an inverse-quantizer which inverse quantizes a value re-quantized by the second quantizer;
re-inverse-quantizer which re-inversely quantizes the value inversely quantized by the inverse-quantizer; and
a subtracter which obtains a difference between the value quantized by the first quantizer and a value equal to N times the value quantized by the second quantizer.
5. The apparatus of claim 4 , further comprising:
a bit plane shifting unit which performs a bit plane shift if a block with a good picture quality is necessary.
6. The apparatus of claim 4 , further comprising:
a maximum value calculating unit which searches a maximum value from output values of the third subtracter; and
a bit plane variable length coding unit which performs variable length coding on the obtained maximum value by bit planes.
7. A fine granularity scalability decoding apparatus, comprising:
a bit plane variable length decoding (VLD) unit which performs variable-length decoding of an enhancement layer bit stream by bit planes;
an adding/subtracting unit which adds/subtracts a difference value between the variable length decoded value of each bit plane and an image reproduced in the base layer;
an inverse-quantizer which inversely quantizes the added/subtracted value;
an inverse discrete cosine transforming unit which restores an image transmitted from the enhancement layer by performing an inverse discrete cosine transforming (IDCT) on the inversely quantized value by block units; and
a clipping unit which restores an enhanced image by clipping the inverse discrete cosine transformed image values in the range of 0~255.
8. The apparatus of claim 7 , wherein the adding/subtracting unit comprising:
a fifth subtracter which obtains a difference value (Δk) between a value equal to N times the value obtained by variable length decoding the bit stream of the base layer and a value obtained by inversely quantizing the variable length decoded value;
a sixth subtracter which obtains a difference value between the value obtained by the fifth subtracter and the bit plane variable length decoded value; and
a seventh subtracter which obtains a difference value between a value obtained by the sixth subtracter and the value inversely quantized by the inverse-quantizer.
9. A fine granularity scalability encoding method, comprising:
quantizing a value obtained by performing a discrete cosine transforming on a motion-compensated image;
re-quantizing the quantized value;
inverse-quantizing the re-quantized value;
obtaining a difference between a value of N times the re-quantized value and the inverse-quantized value;
obtaining a difference between the first-quantized value and the inverse-quantized value; and
subtracting output values obtained in the obtaining steps.
10. The method of claim 9 , further comprising:
finding a maximum value from output values obtained in the subtracting step; and
performing variable length coding on the calculated maximum value by bit planes.
11. A fine granularity scalability encoding method, comprising
quantizing a value obtained by performing a discrete cosine transforming on a motion-compensated image;
re-quantizing the quantized value;
inversely quantizing the re-quantized value;
inversely quantizing the inversely quantized value;
obtaining a difference between the first-quantized value and a value equal to N times the requantized value.
12. The method of claim 11 , further comprising:
finding a maximum value from the output values obtained in the obtaining step; and
performing variable length coding on the calculated maximum value by bit planes.
13. A fine granularity scalability decoding method comprising:
variable-length decoding an enhancement layer bit stream by bit planes;
adding/subtracting a difference value between the variable length decoded value of each bit plane and an image reproduced in a base layer;
inversely quantizing the added/subtracted value;
restoring an image transmitted from the enhancement layer by performing an inverse discrete cosine transforming (IDCT) on the inversely quantized value by block units; and
restoring an enhanced image by clipping the inverse discrete cosine transformed image values in the range of 0~255.
14. The method of claim 13 , wherein the step of adding/subtracting comprises:
obtaining a difference value between a value equal to N times the value obtained by variable length decoding the bit stream of the base layer and the value obtained by inversely quantizing the variable length decoded value;
obtaining a difference value between the value obtained by the first obtaining step and the bit plane variable length decoded value; and
obtaining a difference value between the value obtained in the second obtaining step and the inversely quantized value.