Device for reducing the redundancy in blocks of digital video data in DCT encoding

Claim 2

An improved device for DCT encoding, quantization and variable-word-length encoding of digital video signals with a reduction of redundancy of the video signal, comprising a converter which processes a digital video signal according to a redundancy-reducing algorithm operating on successive blocks of each frame or field and generates a transformed signal which is applied to a quantizer adapted to generate a quantized signal at preset transition levels on opposite sides of zero, including the zero ( . . . , -3, -2, 0, 1, 2, 3, . . . ), which is processed by a processor, adapted to determine the block length and the consecutive zero sequences along a preset scanning order, and adapted to output a symbolic signal, constituted by symbols designating the quantization levels of the non-null data, the zero sequences and a symbol designating the end of the block, the symbolic signal being applied to a variable-length encoder, adapted to encode data according to a preset assignment rule for generating a signal which is applied to a transmission buffer; said quantized signal generated by the quantizer further comprising two ambiguous levels (x1 and x-1) at input values comprised in respective preset intermediate ranges respectively between the zero level and the first positive level and between the zero level and the first negative level, said processor transforming each of said ambiguous levels respectively into a first positive level and a first negative level if the preceding and subsequent input values are both non-null and non-ambiguous levels, and transforms the ambiguous level into zero level in all the other cases.

Claim 3

Device, according to claim 1, wherein said ambiguous levels x1 and x-1 are generated by the quantizer for input values falling respectively within the range between +0.5 and (0.5+k) and within the range between -(0.5+k) and -0.5, k being comprised between 0 and 1 times the interval between the first and second quantization level.

Claim 4

Device, according to claim 2, wherein said ambiguous levels x1 and x-1 are generated by the quantizer also for input values falling respectively within the range between (0.5-h) and +0.5 and within the range between -0.5 and -(0.5-h), h being comprised between 0 and 0.3 times the interval between the zero quantization level and the first quantization level.

Claim 5

A digital video signal compression device comprising: converter means for processing an incoming digital video signal according to a redundancy-reducing algorithm operating on successive blocks of each frame or field to generate a transformed signal; quantizer means operatively connected to said converter means for generating from said transformed signal a quantized signal at preset transition levels including zero and discrete transition levels on opposite sides of zero and further including a first ambiguous level at input values falling into a first preset intermediate range between zero and a preselected positive level and a second ambiguous level at input values falling into a second preset intermediate range between zero and a preselected negative level; processing means operatively connected to said quantizer means for determining block lengths and consecutive zero sequences along a preselected scanning order in said quantized signal, for generating a symbolic signal comprising symbols designating quantization levels of non-zero data, zero sequences and ends of blocks in said quantized signal, and for transforming said first ambiguous level and said second ambiguous level into a first positive level and a first negative level, respectively, if input values immediately preceding and succeeding an instance of one of said first ambiguous level and said second ambiguous level are both non-zero and non-ambiguous levels and for transforming said first ambiguous level and said second ambiguous level into zero in all other cases; variable-length encoder means operatively connected to said processing means for generating an encoded signal encoding data in said symbolic signal according to a predetermined assignment rule; and a transmission buffer operatively connected to said encoder means for receiving and temporarily storing said encoded signal.

Claim 6

A digital video signal compression device comprising: converter means for processing an incoming digital video signal according to a redundancy-reducing algorithm operating on successive blocks of each frame or field to generate a transformed signal; quantizer means operatively connected to said converter means for generating from said transformed signal a quantized signal at preset transition levels including zero and discrete transition levels on opposite sides of zero and further including a first ambiguous level at input values falling into a first preset intermediate range between zero and a preselected positive level and a second ambiguous level at input values falling into a second preset intermediate range between zero and a preselected negative level; processing means operatively connected to said quantizer means for determining block lengths and consecutive zero sequences along a preselected scanning order in said quantized signal, for generating a symbolic signal comprising symbols designating quantization levels of non-zero data, zero sequences and ends of blocks in said quantized signal, and for transforming said first ambiguous level and said second ambiguous level into a first positive level and a first negative level, respectively, if input values immediately preceding and succeeding an instance of one of said first ambiguous level and said second ambiguous level are both non-zero and non-ambiguous levels and for transforming said first ambiguous level and said second ambiguous level into zero in all other cases; variable-length encoder means operatively connected to said processing means for generating an encoded signal encoding data in said symbolic signal according to a predetermined assignment rule; and a transmission buffer operatively connected to said encoder means for receiving and temporarily storing said encoded signal.

Claim 7

The device according to claim 4 wherein said first ambiguous level is generated by said quantizer means for input values falling between +0.5 and (0.5+k) and wherein said second ambiguous level is generated by said quantizer means for input values falling between -(0.5+k) and -0.5, where k is between 0 and 1 times an interval between said first positive level and a second positive level.

Claim 8

The device according to claim 5 wherein said first ambiguous level is also generated by said quantizer means for input values falling between +0.5-h and +0.5 and wherein said second ambiguous level is also generated by said quantizer means for input values falling between -0.5 and -(0.5-h), where h is between 0 and 0.3 times an interval between the zero quantization level and said first positive level.