FR2778039A1 - Method of encoding digital images - Google Patents

Abstract

The signal encoding divides (E2) the signal into blocks (Bp,n). A coding cost function is computed (E5) for each, for a first coding mode and for a scalar quantification coding mode. One coding mode is attributed (E6) to each block, according to cost. The attribution is then iterated to convergence. The cost function used is a weighted sum of the flow and the coding error for the block

Description

The present invention relates to a device and a method for coding a

  digital signal. It also relates to a method and a

  decoding device corresponding to the coding method and device.

  The purpose of coding is to compress the signal, which makes it possible to transmit, respectively memorize, the digital signal by reducing the transmission time, or the transmission rate, respectively by

  reducing the memory space used.

  The invention is in the field of lossy compression

digital signals.

  A known method for coding a digital signal, in this case a digital image, is coding by lattice coded quantization which is described for example in the article entitled "Trellis Coded Quantization of Memoryless and Gauss-Markov Sources" by MW Marcellin and TR Fischer, published in IEEE Transactions on Communications, Vol. 38, n 1, January 1990, as well as in the article "Universal Trellis Coded Quantization" by JH Kasner, MW Marcellin and BR Hunt, available on the Internet at http://vail.ece.arizona.edu/Publications .html. This method has the advantage of minimizing the error of

  quantification, through the use of trellis coding.

  In addition, in order to improve the compression-to-distortion ratio, the inventors have combined this method with another coding mode. In this case, blocks of the signal to be coded are assigned one or the other coding mode, so as to select for each block the most suitable coding mode, that is to say providing the compression ratio. highest distortion. By block, here is meant a set of coefficients extracted from the signal to form a vector. Lattice coded quantization coding applies to a series of blocks. For this coding to be optimal, the sequence must be fixed and known in advance. This means that to optimally assign this mode to a given block, it would first be necessary to know all the blocks to which this mode is assigned, and therefore to know the result of the assignment. There is therefore a difficulty in optimally assigning the coding modes to the blocks and therefore in obtaining a compression-to-distortion ratio.

as high as possible.

  The present invention aims to remedy the drawbacks of the prior art, by providing a device and a compression method

  digital signal offering a high compression-to-distortion ratio.

  To this end, the invention proposes an iterative optimization of the allocation of coding by lattice coded quantization and of another mode of

  coding, to blocks formed in a digital signal.

  More specifically, the invention relates to a digital signal coding method, characterized in that it comprises the steps of: - dividing the signal into blocks, - updating at which a coding cost for each of the blocks is calculated, for at least a first coding mode and for a coding method by lattice coded quantization, - allocation of one of the coding modes to each of the blocks, according to an attribution criterion dependent on the coding cost, - reiteration of the update and allocation steps, so that

  satisfy a convergence criterion.

  Correlatively, the invention proposes a digital signal coding device, characterized in that it comprises: - means for dividing digital signal into blocks, - updating means adapted to calculate a coding cost for each of the blocks, for at least a first coding mode and for a coding mode by lattice coded quantization, means for allocating one of the coding modes to each of the blocks, according to an allocation criterion depending on the cost of coding, the updating and attribution means being adapted to be implemented

  works iteratively, so as to satisfy a convergence criterion.

  The method and the device according to the invention make it possible to obtain a high compression-to-distortion ratio. Indeed, thanks to the iterations of updating and attribution, the invention improves the attribution of the coding modes to

  blocks. The latter are thus coded in a more suitable manner.

  According to a preferred characteristic, the coding cost of each block is a weighted sum of the bit rate and of the coding error of the block considered. This coding cost is simple to implement and gives

satisfactory results.

  According to another preferred characteristic, at a given iteration, the calculation of the coding cost for the coding mode by lattice coded quantization of any block considered, to which the coding mode by lattice coding quantification has been assigned, comprises the stages of: linking of the blocks to which the coding method by lattice coded quantization has been assigned, to form a series of blocks, - coding of the series of blocks by lattice coded quantification of a series of coefficients extracted from the linked blocks,

  - extraction of the bit rate and the coding error of the block considered.

  This calculation makes it possible to precisely determine the cost of coding as

than defined.

  According to a preferred characteristic, at a given iteration, the coding cost for the coding mode by lattice coded quantization of any block considered, to which the first coding mode has been allocated, is the coding cost for the coding mode. coding by lattice coded quantization of the block, calculated during the last iteration during which the mode of

  lattice coded quantization coding was assigned to the block.

  According to a preferred characteristic, the award criterion minimizes the cost of coding. Assigning a coding mode to a block means comparing two coding costs and selecting the associated coding mode

at the lowest cost.

  According to a preferred characteristic, the coding method further comprises an initialization step to which the coding mode by lattice coded quantization is assigned to each of the blocks. This step makes it possible to calculate a first coding cost by quantization coded in

trellis for each of the blocks.

  According to a preferred characteristic, the convergence criterion is satisfied when the same coding modes are respectively allocated to the same blocks in two successive iterations of the allocation step. This

  criterion is reliable while being simple to implement.

  According to a preferred characteristic, an indicator is associated with each block to indicate which coding mode is assigned to each block. This

  flag is then used when decoding the blocks.

  According to a preferred characteristic, the first coding mode is a setting to a predetermined value of the coefficients of the block. This coding mode has the advantages of being a good approximation of blocks containing little information, of being very simple to implement and of having a

  associated coding rate which is zero.

  The invention also relates to a method of decoding a coded digital signal, said signal comprising coded representations of blocks formed in the original signal, each coded representation comprising at least one indicator representative of either a first coding mode or a coding mode by lattice coded quantization, characterized in that it comprises the steps of: - reading the value of the indicator, - decoding the coded representations, as a function of the value of

the respective indicator.

  Correlatively, the invention relates to a device for decoding a coded digital signal, said signal comprising coded representations of blocks formed in the original signal, each coded representation comprising at least one indicator representative of a first coding mode either of a coding method by lattice coded quantization, characterized in that it comprises: - means for reading the value of the indicator, - means for decoding the coded representations, depending

  the value of the respective indicator.

  The decoding method and device make it possible to reconstruct the signal, for example in a receiving device corresponding to a device

  transmitter in which the signal has been coded according to the invention.

  The invention also relates to a digital signal processing apparatus, comprising means for implementing the coding method, or the decoding method, or also comprising the coding device, or the

  decoding device, as set out above.

  The advantages of the coding device, the decoding device and method, of this digital signal processing apparatus, are identical to

  those of the coding method previously exposed.

  An information storage means, readable by a computer or by a microprocessor, integrated or not in the device, possibly removable, stores a program implementing the coding process,

decoding respectively.

  The characteristics and advantages of the present invention will appear more clearly on reading a preferred embodiment illustrated by the attached drawings, in which: - Figure 1 is a block diagram of an embodiment of a device digital signal coding according to the invention; - Figure 2 is a decomposition circuit into frequency sub-bands, included in the device of Figure 1; - Figure 3 is a digital image to be coded by the coding device according to the invention; - Figure 4 is an image broken down into sub-bands by the circuit of Figure 2; - Figure 5 is an image broken down into sub-bands and then divided into blocks; - Figure 6 shows a coding circuit by lattice coded quantization, included in the device of Figure 1; - Figure 7 is a block diagram of an embodiment of a decoding device according to the invention; - Figure 8 is an algorithm for coding a digital signal according to the invention; - Figure 9 is an algorithm for decoding a digital signal

according to the invention.

  According to the embodiment chosen and shown in Figure 1, a coding device according to the invention is intended to code a digital signal in order to compress it. The coding device is integrated into an apparatus 100, which is for example a digital camera, or a digital camcorder, or a database management system, or

still a computer.

  The digital signal to be compressed SI is in this particular mode

  a series of digital samples representing an image.

  The device comprises a signal source 1, here an image signal.

  In general, the signal source either contains the digital signal, and is for example a memory, a hard disk or a CD-ROM, or converts an analog signal into a digital signal, and is for example an analog camcorder associated with a converter analog digital. An output 11 of the signal source is connected to an analysis or sub-band decomposition circuit 2. The decomposition circuit 2 has outputs 21 which are connected to a block division circuit 4. The circuit 4 has outputs 41 connected to

a coding circuit 5.

  The coding circuit 5 comprises several coding circuits, for example a first and a second circuit 51 and 53, which code the blocks received from the circuit 4 according to a first and a second coding mode. The first coding mode here is a setting to a predetermined value of the coefficients of the blocks, and the second coding mode is a lattice coded quantization of a series of symbols extracted from the blocks, called TCQ coding from the English " Coded Quantization Lattice ". The lattice coded quantization is preceded by a link of the blocks to be coded, this link being produced by a link circuit 52 connected at the input of the coding circuit 53. A coding circuit

  entropy 54 is connected at the output of the coding circuit 53.

  An output 51 of circuit 5 is connected to a processing circuit 8, which

  is for example a transmission circuit, or a memory.

  The image source 1 is a device for generating a series of digital samples representing an IM image. The source 1 comprises an image memory and supplies a digital image signal SI to the input of the decomposition circuit 2. The image signal SI is a series of digital words, for example bytes. Each byte value represents one pixel

  of the IM image, here at 256 gray levels, or black and white image.

  The sub-band decomposition circuit 2, or analysis circuit, is a conventional set of filters, respectively associated with decimators in pairs, which filter the image signal in two directions, in high and low sub-bands spatial frequencies. According to FIG. 2, circuit 2 comprises three successive analysis blocks for decomposing the IM image

  in sub-bands according to three levels of resolution.

  In general, the resolution of a signal is the number of samples per unit of length used to represent this signal. In the case of an image signal, the resolution of a sub-band is linked to the number of samples per unit of length used to represent this sub-band horizontally and vertically. The resolution depends on the number of decimations performed, the decimation factor and the resolution of the initial image. The first analysis block receives the digital image signal and applies it to two digital low pass and high pass filters 21 and 22 respectively which filter the image signal in a first direction, for example horizontal in the case an image signal. After passing through decimators by two 210 and 220, the resulting filtered signals are respectively applied to two low-pass filters 23 and 25, and high-pass filters 24 and 26, which filter them in a second direction, for example vertical in the case an image signal. Each resulting filtered signal passes through a decimator with two respective 230, 240, 250 and 260. The first block outputs four sub-bands LL1, LH1, HL1 and HH1 with the highest RES1 resolution

in decomposition.

  The LL1 sub-band comprises the components, or coefficients, of low frequency, in both directions, of the image signal. The LH1 sub-band comprises the low-frequency components according to a first

  direction and high frequency in a second direction, of the image signal.

  The HL1 sub-band comprises the high frequency components in the first direction and the low frequency components in the second direction. Finally, the HH1 sub-band contains the high

  frequency in both directions.

  Each sub-band is an image constructed from the original image, which contains information corresponding to a respectively vertical, horizontal and diagonal orientation of the contours of the image,

  in a given frequency band.

  The LL1 sub-band is analyzed by an analysis block similar to the previous one to provide four LL2, LH2, HL2 and HH2 sub-bands of intermediate RES2 resolution level in the decomposition. The LL2 sub-band comprises the low frequency components according to the two directions of analysis, and is in turn analyzed by the third analysis block similar to the previous two. The third analysis block provides LL3, LH3, HL3 and HH3 sub-bands, with the lowest RES3 resolution in the decomposition,

  resulting from the division into sub-bands of the sub-band LL2.

  Each of the resolution sub-bands RES2 and RES3 also corresponds to an orientation in the image.

  The decomposition performed by circuit 2 is such that a sub-

  band of a given resolution is cut into four sub-bands of lower resolution and therefore has four times more coefficients than each of the

  lower resolution sub-bands.

  A digital IM image at the output of image source 1 is represented schematically in FIG. 3, while FIG. 4 represents the IMD image resulting from the decomposition of the IM image, into ten sub-bands according to three levels of resolution, by circuit 2. The IMD image contains as much information as the original IM image, but the information is

  frequently divided according to three levels of resolution.

  The lower resolution level RES3 comprises the sub-bands LL3, HL3, LH3 and HH3, that is to say the low-frequency sub-bands according to the two directions of analysis. The second resolution level RES2 comprises the sub-bands HL2, LH2 and HH2 and the higher resolution level RES1

  includes the higher frequency sub-bands HL1, LH1 and HH1.

  The different sub-bands of the LL3 sub-band are sub-

retail tapes.

  Of course, the number of levels of resolution, and consequently of sub-bands, can be chosen differently, for example 13 sub-bands and four levels of resolution, for a two-dimensional signal such as an image. The number of sub-bands per resolution level can also be different. The analysis and synthesis circuits are adapted to the

dimension of the signal processed.

  The sub-bands formed by the circuit 2 are supplied to the division circuit 4, according to an a priori arbitrary but predetermined sub-band order. It should be noted that the decomposition into frequency sub-bands of the signal is not essential for the invention, which can be implemented for

  a digital signal that has not been broken down into sub-bands.

  It should also be noted that a sub-band of the digital signal is itself a digital signal, and that according to the invention, it is processed

  independently of the other sub-bands.

  As shown in Figure 5, the division circuit 4 divides each detail sub-band into a block. According to the embodiment chosen, all the sub-bands supplied to circuit 4 are divided into the same number N of blocks Bp ,, where the index p is an integer, here between 1 and 9, which represents the order of the sub-band considered and the index n, between 1 and N, is an integer which represents the order of the block in the sub-band considered. The blocks are here square, but may alternatively be rectangular. Of

  generally, a block is a set of coefficients extracted from the sub-

strip to form a vector.

  The order of the blocks is a priori arbitrary, but predetermined. For practical reasons, blocks are ordered in the same way in

  all sub-bands, for example from left to right and from top to bottom.

  As a consequence of the mode of division into blocks, the surface of the blocks is divided by four while passing from the resolution RES1 to the resolution RES2, and

  from resolution RES2 to resolution RES3.

  This division is simple to implement, since all the subbands are divided into the same number of blocks. However, for the implementation of the invention, the number and the format of the blocks can be

  different from one resolution to another and from one sub-band to another.

  The coding circuit 5 codes each block Bp, n supplied by the circuit 4. To this end, the coding circuit selects the first or the second mode of

  coding, as detailed below.

  The first coding mode (circuit 51) consists in setting all the coefficients of the block to a predetermined value, for example the value zero. This coding is very economical in number of bits, since it does not involve the transmission or storage of any coding data, and therefore requires a reduced transmission rate. However, the error of

  coding is likely to be large if the block considered is not of low energy.

  As shown in FIG. 6, the second coding mode is a lattice coded quantization. The link circuit 52 receives the blocks which are to be coded by lattice coded quantization and links these blocks so as to form a series of blocks {Bm} l, m being an integer between 1 and M, and M being the number of blocks of the suite. To form the sequence, the blocks are for example considered from left to right and from top to bottom, in each frequency band. The

  circuit 52 provides this continuation to coding circuit 53.

  The circuit 53 comprises a coding circuit 531 according to the Viterbi algorithm, a shift register 532, a dictionary selection circuit 533 and memory means 534 for storing vector dictionaries of

coded.

  The coding by lattice coded quantization is described for example

  in the article titled "Trellis Coded Quantization of Memoryless and Gauss-

  Markov Sources "by MW Marcellin and TR Fischer, published in IEEE Transactions on Communications, Vol. 38, no 1, January 1990, as well as in the article" Universal Trellis Coded Quantization by JH Kasner, MW Marcellin and BR Hunt, available by Internet at http://vail.ece.arizona.edu/Publications.html. In general, the coding circuit 53 codes a sequence of symbols {Sk} to provide two bit streams i (k) and j (k), where i (k) represents a sequence of transitions and j (k) represents a sequence of indices of code vector dictionaries. In the context of the invention, the symbols sk are the coefficients extracted from blocks Bp, n, possibly quantified, provided by the

circuit 4.

  The operation of circuit 53 is that of a finite state machine, the transition from one state to another being identified by a transition. In a first embodiment, each state corresponds to a dictionary and is

  identified by the two binary values i (k-2) and i (k-1).

  Each of the dictionaries contains code vectors which are each identified by an index in the dictionary concerned. A code vector is therefore completely identified by its index and by the state

  representing the dictionary to which it belongs.

  The possible transitions of the finite state machine, for the series of symbols to be coded, form a regular structure, or trellis. Circuit 531 implements a Viterbi algorithm to determine an optimal path in the

trellis.

  The path is optimal in the sense of a cost which is minimized over the entire trellis, therefore over the entire sequence to be coded. The cost of a transition is the quadratic error measured between the symbol to be coded and the code vector selected in the dictionary identified by the state in which the transition ends. The cost of a trellis state is the sum of the costs of the transitions leading to this state. The Viterbi algorithm calculates the minimum cost of each state to determine the optimal path represented subsequently by

transitions i (k).

  In a second embodiment, the number of states is greater than the number of dictionaries. For example, four dictionaries and eight states are used. The shift register then stores three binary values for

  define the eight states. Each state is associated with two dictionaries.

  In all cases, the circuit 53 supplies the two bit streams i (k) and j (k) to the entropy coding circuit 54 which combines them and performs coding.

entropy of the combination.

  An indicator Ipn is associated with each of the blocks to indicate which is the coding mode selected by the circuit 5. The indicator Ip, is for example a word whose value represents the selected coding. The selection of the coding mode is carried out by iteration and is detailed below.

(figure 8).

  The circuit 5 transmits to the processing circuit 8 the indicators Ip, n of

  each of the coded blocks, associated with the coded sequence of blocks.

  With reference to FIG. 7, the decoding device globally performs operations reverse from those of the coding device. The decoding device is integrated into an apparatus 200, which is for example a digital image player, or digital video sequence player, or a system

  database management, or even a computer.

  The same device can include both the coding device and the decoding device according to the invention, so as to perform

  coding and decoding operations.

  The decoding device comprises a source of coded data which comprises for example a reception circuit associated with a buffer memory. An output 20 of circuit 20 is connected to a reading circuit 21

  indicator Ip, n, one output 214 of which is connected to a decoding circuit 22.

  The decoding circuit 22 has an output 221 connected to a reconstruction circuit 23. The latter has an output 231 connected to a circuit 24 for processing the decoded data, comprising for example means for

image viewing.

  Circuit 20 provides coded data to circuit 21, which determines

  the coding mode used for each of the blocks by analyzing the indicator Ip, n.

  If the indicator Ip, n indicates that the block considered is coded by zeroing, its decoding consists in creating a block whose all the coefficients are at the value zero. The size of the block created depends on the sub-band being

decoding.

  If the indicator Ipn indicates that the block considered is coded by lattice coded quantization, the circuit 22 globally performs operations opposite to those performed in coding. For each symbol to be decoded, the transition i (k) is decoded to determine a dictionary of code vectors and the index j (k) is decoded to determine a code vector in this

  dictionary. The set of decoded symbols forms a decoded block.

  The circuit 22 supplies the decoded blocks Bdp ,, to the reconstruction circuit 23, which is a synthesis circuit corresponding to the analysis circuit 2 described above and reconstructs the image IMd corresponding to the

decoded sub-bands.

  According to a preferred embodiment of the invention, the sub-band decomposition circuits 2, block division 4, coding 5 and processing 8, all included in the coding device represented in FIG. 1,

  are produced by a microprocessor associated with living and read-only memories.

  The read only memory comprises a program according to the invention for coding each of the data blocks, and the random access memory comprises registers adapted to

  save variables changed during program execution.

  The coding program can be stored in whole or in part in any information storage means capable of cooperating with the microprocessor. This storage means can be read by a computer or by a microprocessor. This storage means is integrated or not to the device, and can be removable. For example, it may include a magnetic strip, a

  floppy disk, or a CR-ROM (compact disk with frozen memory).

  Similarly, the reading 21, decoding 22, reconstruction 23 circuits, included in the decoding device shown in FIG. 6, are produced by a second microprocessor associated with living and read only memories. The read-only memory comprises a program according to the invention for decoding each of the data blocks, and the random access memory comprises registers suitable for recording variables modified during execution

from the program.

  The decoding program can be stored in whole or in part in any information storage means capable of cooperating with the microprocessor. This storage means can be read by a computer or by a microprocessor. This storage means is integrated or not to the device, and can be removable. For example, it may include a magnetic strip, a

  floppy disk, or a CR-ROM (compact disk with frozen memory).

  With reference to FIG. 8, a method of coding according to the invention of an IM image, implemented in the coding device, comprises steps E1 to E9. The coding process uses two coding modes

  attributable to blocks according to a criterion.

  Step E1 is the decomposition into sub-bands of the image IM, as shown in FIG. 4. Step E1 results in the sub-bands LL3, HL3, LH3 and HH3 of lower resolution RES3, the sub -bands LH2, HL2, HH2 of intermediate resolution RES2, and the sub-bands LH1, HL1 and HH1 of higher resolution RES1. As explained above, the decomposition into

  sub-band is not essential for the invention.

  Step E1 is followed by step E2 which is the division of the sub-

  strips in Bpn blocks, as shown in Figure 5.

  The next step E3 is an initialization to consider the first sub-band. The sub-bands are taken into account in any order a priori, while being predetermined. Each sub-band is processed

  independently of the other sub-bands.

  The next step E4 is an initialization at which the indicator Ipn of each of the blocks Bp.n of the current sub-band is set to a value indicating that the block is coded by the second coding mode, namely by

lattice coded quantization.

  Step E4 is followed by step E5 which is the updating of flow rates

  and distortions, for each of the blocks of the sub-band considered.

  At this stage, the following are calculated and stored: - the bit rate RIp, n associated with the coding of the block Bpn by zeroing, - the distortion D1, p ,, associated with the coding of the block Bpn by zeroing, - the bit rate R2, pn associated with the coding of the block Bpn by lattice coded quantization, and - the distortion D2, pn associated with the coding of the block Bp, by

lattice coded quantization.

  The flow R1, p., Associated with the coding of the block Bp, by zeroing is zero. The distortion D1lpn associated with the coding of the block Bpn by zeroing is equal to the quadratic error of the block. These two quantities are calculated only once, during the first passage through step E5. These quantities do not

  not vary during subsequent iterations of coding.

  To determine the rate R2, p, n and the distortion D2, p, n associated with the coding of the block Bp, n by quantization coded in trellis, it is necessary to consider two cases. In the first case, the indicator Ip, n of the block Bp, n indicates that the block

  is to be coded by lattice coded quantization.

  The blocks which are to be coded by lattice coded quantization are linked so as to form a series of blocks to be coded {B4,}. The blocks are for example considered from left to right and from top to bottom, in the frequency band. The block Bp, n then forms part of this series of blocks to be coded by lattice coded quantization. The coding of the block sequence is carried out, as previously explained (FIG. 6). At the first iteration, all the blocks of the sub-band considered are part of the sequence to be coded by quantization

trellis coded.

  The coding rate of the block Bp ,, is determined. For this, the entropy of the coded sequence is determined and the contribution of the block considered is extracted. The coding distortion of the block Bp ,, is determined relative to the

original block.

  The error D2, pn measures the quadratic error induced in the image reconstructed by the coding of the block considered. In the case where the subband decomposition is orthogonal, the error D2, pn is equal to the quadratic error

  between the original block and the reconstructed block.

  In the second case, the indicator Ipn of the block Bpn indicates that the block is to be coded by the first coding mode, ie by zeroing. It is not part of the series of blocks to be coded by lattice coded quantization. The coding rate and distortion by the second coding mode (trellis coded quantization) are estimated by their respective value determined and stored during the last iteration during which the block considered

  was in a series of blocks to be coded by lattice coded quantization.

  The next step E6 is the allocation, for each of the blocks of the subband being processed, of the coding mode which is most suitable for it, according to a criterion. The criterion consists in minimizing a cost of

  coding, which is generally a function of bit rate and distortion.

  According to a preferred embodiment, the coding cost is the sum Rj. pn + o.D, p, n, o X, is a coefficient for adjusting the compression / distortion ratio, and j is an integer equal to 1 or 2, representing the first

or the second coding mode.

  The coefficient X varies from zero to infinity. For practical reasons, the sum (1-X) can be used in an equivalent manner. Rjp, n + X.Djpn, with the

  coefficient X varying between zero and one.

  The coding mode which minimizes the sum Rj, p, n +; Djpn is selected to code the block considered Bpn Consequently, the value of

  the indicator Ip, n of the block considered is updated.

  The next step E7 is a convergence test. This test consists in checking whether the same coding modes are allocated to the same blocks, during several, for example two, successive repetitions of steps E5 and E6. As long as the convergence test is not satisfied, step E7 is followed by step E5. When the convergence test is satisfied, step E7 is

followed by step E8.

  As a variant, a maximum number of repetitions of steps E5 and E6

  is predetermined, so as to limit the coding time.

  The coding result of the sub-band considered is a set of indicators indicating for each block which coding mode has been allocated to it after convergence, and the rest of the data coded by lattice coded quantization.

  The next step E8 is a test to determine if all the sub-

  tapes were processed. If at least one sub-band remains to be processed, step E8 is followed by step E9 to consider the next sub-band. Step E9

  is followed by step E4 previously described.

  With reference to FIG. 9, a method of decoding according to the invention of an IM image, implemented in the decoding device, comprises steps E21 to E30.

  Step E21 is an initialization to consider the first sub-

strip to decode.

  Step E21 is followed by step E22 which is an initialization for

  consider the first block to decode in the current sub-band. Money-

  bands are decoded in the same order as coding, and blocks in a given subband are decoded in the same order as coding, well

  that different orders are possible.

  The next step E23 is the reading of the indicator Ipn to determine

  which coding mode was used to code the current block.

  Step E23 is followed by step E24 which is the decoding of the current block. If the block has been coded by zeroing, decoding consists in creating a block whose all the coefficients are at zero. The size of the block created depends on the sub-band being decoded, and is for example determined by the index of the block. If the block has been coded by lattice coded quantization, operations opposite to those performed in coding are performed. The coding data of the block in question is extracted from the sequences of indices and transitions. For each symbol to be decoded, the transition i (k) is decoded to determine a dictionary of code vectors and the index j (k) is decoded to determine a code vector in this dictionary. All

  decoded symbols form a decoded block.

  The decoded block Bdp.n is stored in the next step E25.

  The steps E26 and E28 are tests to check, respectively if all the blocks of a sub-band, and if all the sub-bands have been decoded. If at least one block remains to be decoded in the current sub-band, step E26 is followed by step E27 to consider the next block. The stage

  E27 is followed by step E23 previously described.

  If at least one sub-band remains to be decoded, step E28 is followed by step E29 to consider the next sub-band. Step E29 is followed

  by step E22 previously described.

  When all the sub-bands have been decoded, that is to say that the response is positive in step E28, this last step is followed by step E30 of construction of the decoded image. The latter can then be

visualized, for example.

  Of course, the present invention is in no way limited to the embodiments described and shown, but includes, on the contrary,

  any variant within the reach of the skilled person.

  In particular, the invention can easily be applied to other

types of signals.

  These signals can be mono-dimensional signals such as sounds, or seismic readings, or even electrocardiograms; depending on their nature, the signals are analyzed according to frequencies

temporal or spatial.

  These signals can be three-dimensional such as video sequences represented according to two spatial frequencies and one temporal frequency. We then implement a decomposition into frequency sub-bands in three dimensions, and the decomposition of the signal into vectors

  is also done in dimension three.

  For a signal with components in multiple frequency bands, such as a color image signal with red components,

  green and blue, the invention applies in each of the frequency bands.

Claims (40)

  1. Method for coding a digital image signal, characterized in that it comprises the steps of: - dividing (E2) the signal into blocks (Bp.), - updating (E5) at which a cost is calculated coding of each of the blocks, for at least a first coding mode and for a coding mode by lattice coded quantization, - allocation (E6) of one of the coding modes to each of the blocks, according to a criterion of allocation depending on the cost of coding, - reiteration of the updating and allocation steps, so as to   satisfy a convergence criterion.   2. Coding method according to claim 1, characterized in that the coding cost of each block is a weighted sum of the bit rate (R1, p ,,,   R2p ,,) and the coding error (D1, pn, D2, pn) of the block considered.   3. Coding method according to claim 1 or 2, characterized in that, at a given iteration, the calculation of the coding cost for the coding mode by lattice coded quantization of a block considered to be arbitrary, to which the lattice coded quantization coding mode, comprises the steps (E5) of: - linking the blocks to which the lattice coded quantization coding mode has been assigned, to form a block sequence, - coding the block sequence by lattice coded quantization of a series of coefficients extracted from the linked blocks,   - extraction of the bit rate and the coding error of the block considered.   4. Coding method according to any one of claims 1 to   3, characterized in that, at a given iteration, the coding cost for the coding mode by lattice coded quantization of any block considered, to which the first coding mode has been assigned, is the coding cost for the coding mode by lattice coded quantization of the block, calculated during the last iteration during which the coding mode by   lattice coded quantization was assigned to the block.   5. Coding method according to any one of claims 1 to   4, characterized in that the award criterion minimizes the cost of coding.   6. Coding method according to any one of claims 1 to   , characterized in that it further comprises an initialization step (E4) to which the coding mode by lattice coded quantization is assigned to each of the blocks.   7. Coding method according to any one of claims 1 to   6, characterized in that the convergence criterion is satisfied when the same coding modes are respectively assigned to the same blocks in two   successive iterations of the allocation stage.   8. Coding method according to any one of claims 1 to   7, characterized in that an indicator (Ip, n) is associated with each block (Bpn) for   indicate which coding mode is assigned to each block.   9. Coding method according to any one of claims 1 to   8, characterized in that the first coding mode is a setting to a value   predetermined coefficients of the block.   10. Method for decoding a digital coded image signal, said signal comprising coded representations of blocks formed in the original signal, each coded representation comprising at least one indicator (Ip,) representative of a first mode coding or a coding method by lattice coded quantization, characterized in that it comprises the steps of: - reading (E23) of the value of the indicator (Ip, n), - decoding (E24) of the coded representations, depending on the value of the respective indicator.   11. Digital signal coding device characterized in that it comprises: - means (4) for dividing digital signal into blocks (Bp ,,), - updating means (5) adapted to calculate a cost coding of each of the blocks, for at least a first coding mode and for a coding mode by lattice coded quantization, - means of allocation (E6) of one of the coding modes to each of the blocks, according to an allocation criterion depending on the coding cost, the updating and attribution means being adapted to be implemented   works iteratively, so as to satisfy a convergence criterion.   12. Coding device according to claim 11, characterized in that the updating means (5) are adapted to calculate the coding cost of each block as a weighted sum of the bit rate (R1, p ,, R2, pn) and error   (D1, p, n, D2, p,) coding of the block considered.   13. Coding device according to claim 11 or 12, characterized in that the updating means (5) for calculating the coding cost for the coding mode by lattice coded quantification of a block considered to be arbitrary, to which the coded coded lattice coding mode has been assigned, comprise: - means (52) for connecting the blocks to which the coded coded lattice coding coding mode has been assigned, to form a series of blocks, - means ( 53) for coding the series of blocks by lattice coded quantization of a series of coefficients extracted from the linked blocks, - means for extracting the bit rate and the coding error of the block in question.   14. Coding device according to any one of claims   11 to 13, characterized in that the means (5) for updating, for the calculation, the coding cost for the coding mode by lattice coded quantization of any block considered, to which the first coding mode has been assigned , are adapted to consider the coding cost for the coding mode by lattice coded quantization of the block, calculated during the last iteration during which the coding mode by lattice coded quantization was assigned to the block.   15. Coding device according to any one of claims   11 to 14, characterized in that the allocation means are adapted to put   implement an award criterion which minimizes the cost of coding.   16. Coding device according to any one of the claims   11 to 15, characterized in that the allocation means are adapted to initially assign the coding mode by lattice coded quantization to each of the blocks.   17. Coding device according to any one of claims   11 to 16, characterized in that it is suitable for implementing a convergence criterion which is satisfied when the same coding modes are respectively assigned to the same blocks in two successive iterations of the allocation stage.   18. Coding device according to any one of claims   11 to 17, characterized in that it is adapted to associate an indicator (Ipn) with each block (Bp.n) to indicate which coding mode is assigned to each block.   19. Coding device according to any one of claims   1 to 18, characterized in that it is suitable for implementing a first coding mode which is a setting to a predetermined value of the coefficients of the block.   20. Coding device according to any one of claims   11 to 19, characterized in that the means of division, allocation, quantification, training and selection are incorporated in: - a microprocessor, - a read only memory comprising a program for coding each of the data blocks, and - a random access memory comprising registers adapted to record   variables modified during the execution of said program.   21. Device for decoding a coded digital signal, said signal comprising coded representations of blocks formed in the original signal, each coded representation comprising at least one indicator (Ipn) representative of either a first coding mode or d '' a coding method by lattice coded quantization, characterized in that it comprises: - means for reading (E23) the value of the indicator (Ipn '), - means for decoding (E24) the coded representations , depending on the value of the respective indicator. 22. Decoding device according to claim 21, characterized in that the reading means and the decoding means are incorporated in: - a microprocessor, - a read only memory comprising a program for decoding each of the data blocks, and - a memory alive with registers adapted to record   variables modified during the execution of said program.   23. Apparatus (100) for digital signal processing, characterized in that it includes means suitable for implementing the method of   coding according to any one of claims 1 to 9.   24. Apparatus (200) for digital signal processing, characterized in that it includes means suitable for implementing the method of   decoding according to claim 10.   25. Apparatus (100) for digital signal processing, characterized in that it comprises the coding device according to any one of claims 11 to 20.   26. Apparatus (200) for digital signal processing, characterized in that it comprises the decoding device according to any one of claims 21 or 22.   27. Digital camera characterized in that it comprises means adapted to implement the coding method according to   any one of claims 1 to 9.   28. Digital camera characterized in that it   comprises the coding device according to any one of claims 11   at 20. 29. Digital camcorder characterized in that it comprises means suitable for implementing the coding method according to one   any of claims 1 to 9.   30. Digital camcorder characterized in that it comprises means suitable for implementing the decoding method according to the claim 10.   31. Digital camcorder characterized in that it includes the   coding device according to any one of claims 11 to 20.   32. Digital camcorder characterized in that it includes the   decoding device according to any one of claims 21 or 22.   33. Coding method according to any one of claims 1   to 9, characterized in that it is implemented by a management system of database.   34. Coding method according to claim 10, characterized in that   that it is implemented by a database management system.   35. Coding device according to any one of claims   1 1 to 20, characterized in that it is implemented in a management system database.   36. Decoding device according to any one of   claims 21 or 22, characterized in that it is implemented in a   database management system.   37. Computer characterized in that it includes means suitable for implementing the coding method according to any one of claims 1 to 9.   38. Computer characterized in that it includes suitable means   implementing the decoding method according to claim 10.   39. Computer characterized in that it comprises the device for   coding according to any one of claims 11 to 20.   40. Computer characterized in that it comprises the device for   decoding according to any one of claims 21 or 22.