US4672441A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method and apparatus for picture data reduction for digital video signals
Assignee: Siemens AG
•
Filing: 1986-04-03
•
Granted: 1987-06-09
•
Expires: 2006-04-03
Abstract
Abstract Method and apparatus for the reduction of picture data for digital video signals comprising a processing of the signals by means of block by block transformation method so that a transformed and quantized signal which was generated at a time t-1 and placed in an image storage is subtracted from a transformed signal that occurs at a time t and whereby the difference signal obtained is subject to quantization and the quantized difference signal is subjected to an analysis and to a time delay which corresponds to the time requirement for the analysis for updating the content of the image storage. The signal delayed is added to the signal read out from the image storage which is also delayed and is added dependent on the addition condition signal obtained from the analysis and is subjected to an entropy coding dependent on the analysis results with the addition condition signals containing information as to whether a block which has been analyzed has been concluded is a moved or unmoved block and when the block is a moved block containing information regarding a coefficient group to be transmitted. The signal coded in such fashion is subjected to a buffering and depending on the degree of buffer filling a quantization stage and an analysis stage is influenced so that a signal from a buffer control is supplied to the quantization stage for the purpose of selecting one of a plurality of predetermined quantization characteristics and a second signal is supplied from the buffer control means to the analysis stage to select the maximum number of coefficient groups and a third signal is supplied to the analysis stage from the buffer control for the purpose of deciding whether a block is to be transmitted or not and the coefficients represent the digitized video signal transformed by block which is subdivided into coefficient groups.
Inventors
- 👤 Herbert Hoelzlwimmer
- 👤 Walter Tengler
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-04 04:14
CPC Classifications
H; H04; H04N; H04N19/00; H04N19/10; H04N19/102; H04N19/124; H04N19/132; H04N19/134; H04N19/136; H04N19/137; H04N19/14; H04N19/146; H04N19/149; H04N19/152; H04N19/169; H04N19/17; H04N19/176; H04N19/60; H04N19/61; H04N19/619; H04N19/13; H04N19/50; H04N19/90; H04N19/91
Claims Analysis
Total Claims: 24
Claim 1
Independent Claim
485 words
A method for picture data reduction for digital video signals, comprising a pre-processing of the signals by means of block-by-block transformation method, whereby a transformed and quantized signal that was generated at a time of t-1 and deposited in an image store is subtracted from a transformed signal that occurs at a time t to obtain a difference signal and whereby the difference signal is subjected to a quantization, comprising the steps of, analyzing the quantitized difference signal and delaying for a time (VZ) which corresponds to the time requirement for the analysis (AS) said difference signal for updating the content of the image store, adding delayed signal to the signal read out from the image store (M) which has been correspondingly delayed, said adding being dependent on an addition condition signal acquired from said analysis, and applying an entropy coding (HC) to said signal which depends on the results of said analysis, said addition condition signals containing information as to whether a block whose analysis has been concluded is a "moved"or "unmoved" block and, in case said block is a "moved" block, containing information regarding a coefficient group to be transmitted; buffering the signal coded in such fashion with a buffer (B) for supplying an output signal channel a uniform data flow for transmission, obtaining said uniform data flow from the non-uniform data flow of the entropy coding; influencing depending on the degree of the fill of the buffer a quantization stage (Q) and an analysis stage (AS) whereby a signal from a buffer control means (BC) is supplied to the quantization stage (Q), and selecting one of a plurality of pre-determined quantization characteristics, supplying a second signal from the buffer control means (BC) to the analysis stage (AS) for selecting the maximum number of coefficient groups, and supplying a third signal to the analysis stage (AS) from the buffer control means (BC) for deciding whether a block is to be transmitted or is not to be transmitted; subdividing the coefficients representing the digitized video signal transformed block-by-block into coefficient groups according to prescribed rules; and identifying a value for each of the coefficient groups in a calculation stage (E), and determining from said value whether a super-group formed in a decision means (S) from neighboring coefficient groups is to be transmitted is selected, whereby coefficient groups which are not to be transmitted based on said identified value can also be arranged in such a supergroup, and secondly, classifying in a mediated fashion, following a step-by-step summation E.sub.I (i)=E.sub.I (i-1)+E(i) of all values respectively belonging to a block in an integrator (I), whereby i=2 . . . 31 preferably applies, whereby E(i) is the value for the coefficient group i and whereby E I (1)=E(1) applies, said classification serving the purpose to decide whether a block is to be transmitted and the manner in which a block to be transmitted is to be coded.
Claim 2
Independent Claim
485 words
A method for picture data reduction for digital video signals, comprising a pre-processing of the signals by means of block-by-block transformation method, whereby a transformed and quantized signal that was generated at a time of t-1 and deposited in an image store is subtracted from a transformed signal that occurs at a time t to obtain a difference signal and whereby the difference signal is subjected to a quantization, comprising the steps of, analyzing the quantitized difference signal and delaying for a time (VZ) which corresponds to the time requirement for the analysis (AS) said difference signal for updating the content of the image store, adding delayed signal to the signal read out from the image store (M) which has been correspondingly delayed, said adding being dependent on an addition condition signal acquired from said analysis, and applying an entropy coding (HC) to said signal which depends on the results of said analysis, said addition condition signals containing information as to whether a block whose analysis has been concluded is a "moved"or "unmoved" block and, in case said block is a "moved" block, containing information regarding a coefficient group to be transmitted; buffering the signal coded in such fashion with a buffer (B) for supplying an output signal channel a uniform data flow for transmission, obtaining said uniform data flow from the non-uniform data flow of the entropy coding; influencing depending on the degree of the fill of the buffer a quantization stage (Q) and an analysis stage (AS) whereby a signal from a buffer control means (BC) is supplied to the quantization stage (Q), and selecting one of a plurality of pre-determined quantization characteristics, supplying a second signal from the buffer control means (BC) to the analysis stage (AS) for selecting the maximum number of coefficient groups, and supplying a third signal to the analysis stage (AS) from the buffer control means (BC) for deciding whether a block is to be transmitted or is not to be transmitted; subdividing the coefficients representing the digitized video signal transformed block-by-block into coefficient groups according to prescribed rules; and identifying a value for each of the coefficient groups in a calculation stage (E), and determining from said value whether a super-group formed in a decision means (S) from neighboring coefficient groups is to be transmitted is selected, whereby coefficient groups which are not to be transmitted based on said identified value can also be arranged in such a supergroup, and secondly, classifying in a mediated fashion, following a step-by-step summation E.sub.I (i)=E.sub.I (i-1)+E(i) of all values respectively belonging to a block in an integrator (I), whereby i=2 . . . 31 preferably applies, whereby E(i) is the value for the coefficient group i and whereby E I (1)=E(1) applies, said classification serving the purpose to decide whether a block is to be transmitted and the manner in which a block to be transmitted is to be coded.
Claim 3
Dependent Claim
53 words
A method according to claim 1, wherein the coefficient groups are formed such that coefficients (y(u, v)) whose matrix indices (u,v) meet the condition u+v=i-1 are respectively combined to form a coefficient group (i), whereby u, v=0 . . . 15 and whereby u, v are the horizontal or, respectively, vertical discrete frequencies.
Claim 4
Independent Claim
63 words
A method according to claims 1 or 2, characterized in that said value for the selection of a coefficient group and for classification is ##EQU6## whereby i is the number of a coefficient group, whereby k can have the values 1, 2, 3 . . . , and whereby Δy Q (u, i-1-u) is the quantized difference signal of the coefficients (u, i-1-u).
Claim 5
Dependent Claim
45 words
A method according to claim 3, characterized in that the value for the selection of a coefficient group and for the classification is the energy ##EQU7## of a coefficient group (i), whereby Δy Q (u, i-1-u) is the quantized difference signal of the coefficients (u,i-1-u).
Claim 6
Dependent Claim
50 words
A method according to claim 3, characterized in that the value for the selection of a coefficient group and for the classification is the sum of the absolute values ##EQU8## of a coefficient group (i), whereby Δy Q (u, i-1-u) is the quantized difference signal of the coefficients (u, i-1-u).
Claim 7
Dependent Claim
37 words
A method according to claim 1, characterized in that the quantization is executed such that respectively like quantization intervals are employed for all amplitude ranges of the signal to be quantized, so that a linear quantization results.
Claim 8
Dependent Claim
36 words
A method according to claim 6, characterized in that the quantization is executed dependent on the picture activity; and a selection of one of four quantization intervals which respectively differ by the factor "2" is accomplished.
Claim 9
Dependent Claim
42 words
A method according to claim 1, characterized in that a block-wise classification in four classes is carried out, said classes serving the purpose of respectively selecting one of four coding allocation tables, whereby one of these four classes is an "unmoved" class.
Claim 10
Dependent Claim
160 words
A method according to claim 8, characterized in that the classification is executed such that, in a first step, a decision is made from a data field respectively supplied by the integrator (I) as to whether the appertaining block is to be classified as "moved" or "unmoved", whereby, when the block is classified as "unmoved", a 2-bit code word is generated for said class 4 which is the "unmoved" class or, respectively, a decision is made in a second and, under given conditions, a third step concerning which of three "moved" classes, namely class 1 . . . class 3, said block is to be allocated to, whereby the energy belonging to the supergroup to be transmitted is identified from the data field (E I ) and is successively compared to two thresholds read out and edited from first and second tables, a 2-bit code word being respectively generated under given conditions for said classes 1 . . . 3.
Claim 11
Dependent Claim
75 words
A method according to claim 1, characterized in that seven code tables of variable word length are employed for the purpose of coding, whereby on code table (1) is provided for extremely small signal variances and one code table (7) is provided for extremely large signal variances, whereby the code table (i) respectively differs from the code table (i+1) in that it is generated for a signal variance which is greater by the factor "4".
Claim 12
Independent Claim
50 words
A method according to claims 9 or 10, characterized in that the allocation of the code tables to the individual coefficients of a block is done with the assistance of three allocation tables of which one is selected by means of one of the identified classes for a "moved" block.
Claim 13
Dependent Claim
66 words
A method according to claim 1, characterized in that an eighth code table is provided for coding the supergroup to be transmitted; and an auxiliary table is provided which is employed to select therefrom the code word number of the eighth code table allocated to this supergroup for every combination of the values (N O and N D ) for limiting the super-group to be transmitted.
Claim 14
Dependent Claim
17 words
A method according to claim 10, characterized in that the seven code tables are Huffman code tables.
Claim 15
Dependent Claim
16 words
A method according to claim 12 wherein the eighth code table is a Huffman code table.
Claim 16
Dependent Claim
47 words
A method according to claim 1 characterized in that a buffer control means (BC) defines the maximum allowable number of coefficient groups as well as the threshold for the decision "moved"/"unmoved" and modifies the quantization resolution "memory-free" dependent on the degree that the buffer memory is filled.
Claim 17
Dependent Claim
184 words
A method according to claim 15, characterized in that the maximum number of coefficient groups (N NMAX ) is defined by the relationship ##EQU9## wherein trun (truncated) is the "cut-off" function of the illustrated function and whereby B N preferably lies in the range of 0.3 . . . 0.6; and in that the threshold (T) for the decision "moved"/"unmoved" is varied by the relationship ##EQU10## wherein B T preferably lies in the range of 0.6 . . . 0.8; and in the quantizer resolution Δ is varied with an intermediate value Δ * which varies with the degree of the filling of the buffer memory and is related to the buffer memory length, which varies by the relationship ##EQU11## whereby the quantizer resolution is established from this intermediate value Δ * under the following conditions: when 0.75Δ o ≦Δ* applies, then Δ=Δ o applies, when 0.375Δ o ≦Δ*<0.75Δ o , then Δ=Δ o /2 applies, when 0.1875Δ o ≦Δ*<0.375Δ o applies, then Δ=Δ o /4 applies, when Δ*<0.1875Δ o applies, thenΔ=Δ o /8 applies, whereby B.sub.Δ preferably has values between 0.4 and 0.6.
Claim 18
Dependent Claim
44 words
A method according to claim 16, characterized in the tabular values are determined for the relationships for N DMAX , T and Δ, and these tabular values are addressed on the basis of the six most significant bits of the degree of buffer filling.
Claim 19
Dependent Claim
343 words
Apparatus for picture data reduction for digital video signals comprising, a receiver with a receiver buffer memory (B E ), a decoder (DC) connected to said receiver buffer memory B E , a reconstruction means (R) connected to said decoder, a receiver buffer control means (BC E ) connected between said receiver buffer memory B E and said reconstruction means (R), a receiver summing element (+ E ) receiving inputs from said decoder and said reconstruction means, and a receiver image store (M E ) connected to said receiver summing element, the data-reduced digital video signal is supplied to said receiver buffer memory (B E ) after a channel decoding; said decoder (DC) reconstructs a signal with constant word length from the preferably Huffman-coded signal intermediately stored in the receiver buffer memory (B E ); the reconstruction means (R) reproduce representative values from the numbers for representative values coded with constant word length and from a signal supplied by the receiver buffer control means (BC E ) for the selection of one of a plurality of quantizer tables, said receiver buffer control means (BC E ) produces the output signal as the sourceside quantization signal, and said receiver buffer control means receives a signal indicative of the degree of filling (B) of the buffer memory from the receiver buffer memory (B E ); the representative values from the reconstruction means (R) are supplied to the receiver summing element (+ E ), and addition condition signals (N O ,N D and a signal for "moved"/"unmoved") from the decoder (DC) as well as the picture signal from the receiver image store (M E ) reconstructed at time t-1 are also supplied, and the reconstructed difference signal at a time t from the reconstruction means (R) is added to the picture signal reconstructed at a time t-1, with the addition accomplished depending on the addition condition signals; and the reconstructed transformed picture signal obtained by said addition supplied to a inverse transformation stage (IT) and, also is supplied to said receiver image store (M E ).
Claim 20
Dependent Claim
343 words
Apparatus for picture data reduction for digital video signals comprising, a receiver with a receiver buffer memory (B E ), a decoder (DC) connected to said receiver buffer memory B E , a reconstruction means (R) connected to said decoder, a receiver buffer control means (BC E ) connected between said receiver buffer memory B E and said reconstruction means (R), a receiver summing element (+ E ) receiving inputs from said decoder and said reconstruction means, and a receiver image store (M E ) connected to said receiver summing element, the data-reduced digital video signal is supplied to said receiver buffer memory (B E ) after a channel decoding; said decoder (DC) reconstructs a signal with constant word length from the preferably Huffman-coded signal intermediately stored in the receiver buffer memory (B E ); the reconstruction means (R) reproduce representative values from the numbers for representative values coded with constant word length and from a signal supplied by the receiver buffer control means (BC E ) for the selection of one of a plurality of quantizer tables, said receiver buffer control means (BC E ) produces the output signal as the sourceside quantization signal, and said receiver buffer control means receives a signal indicative of the degree of filling (B) of the buffer memory from the receiver buffer memory (B E ); the representative values from the reconstruction means (R) are supplied to the receiver summing element (+ E ), and addition condition signals (N O ,N D and a signal for "moved"/"unmoved") from the decoder (DC) as well as the picture signal from the receiver image store (M E ) reconstructed at time t-1 are also supplied, and the reconstructed difference signal at a time t from the reconstruction means (R) is added to the picture signal reconstructed at a time t-1, with the addition accomplished depending on the addition condition signals; and the reconstructed transformed picture signal obtained by said addition supplied to a inverse transformation stage (IT) and, also is supplied to said receiver image store (M E ).
Claim 21
Dependent Claim
22 words
Apparatus according to claim 18, characterized in that a counter is provided which measures the degree of filling of said buffer memory.
Claim 22
Dependent Claim
21 words
Apparatus according to claim 18, characterized in that a shift register is provided for delaying by n blocks the video signals.
Claim 23
Dependent Claim
21 words
Apparatus according to claim 18, characterized in that a gate chain is provided for delaying by n blocks said video data.
Claim 24
Dependent Claim
18 words
Apparatus according to claim 18, characterized in that at least one ROM memory is used for storing tables.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1990-06-08 | 1990-12-08 | 1991-06-09 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1994-06-08 | 1994-12-08 | 1995-06-09 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1998-06-08 | 1998-12-08 | 1999-06-09 | $7700.00 | Window Closed / Lapsed |