WO2016097556A1

WO2016097556A1 - Method for encoding a digital image, decoding method, devices, user terminal and computer programs for same

Info

Publication number: WO2016097556A1
Application number: PCT/FR2015/053470
Authority: WO
Inventors: Félix Henry
Original assignee: B<>Com; Orange
Priority date: 2014-12-19
Filing date: 2015-12-14
Publication date: 2016-06-23
Also published as: FR3030975B1; FR3030975A1

Abstract

The invention relates to a method for encoding a digital image, said image being divided into a plurality of blocks of pixels processed in a predefined order, said method including the following steps, performed for a current block (C): processing (E1) the current block intended to provide a set of description elements of the processed block; selecting (E2) a subset of description elements to be predicted from the provided set; predicting (E4) description elements of the selected subset; encoding (E3) values of the non-predicted description elements; calculating (E5) a prediction indicator (IP) for the description elements selected from predictions of the selected description elements and the original values thereof; encoding (E6) the resulting values of indicators for the predicted description elements. The method is characterized in that the selection step (E2) seeks, for a subset of the set of description elements, an order for scanning the description elements of the subset for which a prediction of the current element adopts the same value for all the possible values of the other elements of the subset that remain to be scanned and selects the subset, when a scanning order has been found.

Description

Method of encoding a digital image, decoding method, devices, user terminal and associated computer programs

1. Field of the invention

The field of the invention is that of signal compression, in particular of a digital image or of a sequence of digital images, in which a prediction of a portion of the signal to be coded is made from a portion of the already coded signal.

The encoding / decoding of digital images applies in particular to images from at least one video sequence comprising:

images coming from the same camera and succeeding each other temporally (coding / decoding of 2D type),

images from different cameras oriented according to different views (coding / decoding of 3D type),

corresponding texture and depth components (3D type coding / decoding),

- etc.

The present invention applies similarly to the coding / decoding of 2D or 3D type images.

The invention may especially, but not exclusively, apply to video coding implemented in current AVC and HEVC video encoders and their extensions (MVC, 3D-AVC, MV-HEVC, 3D-HEVC, etc.), and to corresponding decoding.

The invention can also be applied to audio coding, for example implemented in current audio encoders (EVS, OPUS, MPEG-H, etc.) and their extensions and the corresponding decoding.

2. Presentation of the prior art

We consider a conventional compression scheme of a digital image, according to which the image is divided into blocks of pixels. A current block to be coded is predicted from a previously coded decoded block. A residual block is obtained by subtracting the original values from the predicted values. It is then transformed using a DCT type transformation (for "Discrete Cosine Transform", in English) or wavelets. The transformed coefficients are quantized and their amplitudes are subjected to entropic coding of the Huffmann or arithmetic type. Such coding obtains effective performances because, because of the transformation, the values of the amplitudes to be coded are for the most part zero. On the other hand, it does not apply to the values of the signs of the coefficients, whose + and - values are generally associated with equivalent appearance probabilities. Thus, the signs of the coefficients are coded by a bit 0 or 1.

The article by Koyama, J. et al, titled "Coefficient signal bit compression in video coding", and published in the proceedings of the "Picture Coding Symposium (PCS)" conference in May 2012, is known from a selection method. of coefficient signs of a residual block to predict. The proposed selection is based on a predetermined number of coefficients as a function of their amplitude and the size of the block from which they come. The selected signs are predicted by testing all the possible combinations of the values of these signs for the block and choosing the one that maximizes a predetermined likelihood criterion. The predictions obtained are compared with the original values of the signs to determine the value of a prediction indicator. , also called residue of a predicted sign. This indicator can take two values, which are a first value representative of a correct prediction and a second value representative of an incorrect prediction. The rest of the signs are coded explicitly, without prediction.

An advantage of such a selection is to predict the value of a sign with a correct prediction probability greater than 50%, thus to allow the application of entropic coding to the values of the prediction indicators. This entropic coding encodes the sign information with an average bit rate less than one bit per sign, and thus makes it possible to increase the compression ratio.

3. Disadvantages of prior art

Nevertheless, this technique has at least two major disadvantages:

- Some signs whose correct prediction probability is close to 50% enter the selection of coefficients to predict. If this has no impact on the compression performance (no gain), there is an unnecessary increase in the number of calculations to be performed; - Some coefficients whose probability of correct prediction is high (greater than 50%) are not retained in the selection of coefficients to be predicted. There is then a loss of compression efficiency, because these coefficients could be used to further reduce the size of the coded signal.

4. Objectives of the invention

The invention particularly aims to overcome these disadvantages of the prior art.

More specifically, an object of the invention is to propose a solution that more effectively selects the signs to be predicted.

Another objective of the invention is to propose a solution that is more efficient in compression.

Yet another object of the invention is to propose a solution that applies to any type of description element of a block of pixels used for coding a digital image.

5. Presentation of the invention

These objectives, as well as others that will appear later, are achieved by a method of coding a digital image, said image being divided into a plurality of blocks of pixels processed in a defined order, said method comprising the following steps, implemented for a current block:

- Processing of the current block for providing a set of description elements of the processed block;

Selecting a subset of description elements to be predicted from the provided set;

Prediction of the description elements of the selected subset;

Coding of the values of the unpredicted description elements;

Calculation of a prediction indicator of the selected description elements from the predictions of the selected description elements and their values original, the indicator being intended to take a value in a group comprising:

a first value representative of a correct prediction;

a second value representative of an incorrect prediction;

- Coding of the indicator values obtained for the predicted elements of description;

According to the invention, said method is particular in that the selection step searches, for a subset of the set of description elements, of a search order of the description elements of the subset for which an element current is predicted with the same value for all possible values of the remaining subset elements to be scanned and selects the subset, when a browse order has been found.

The invention generalizes the fact of selecting the signs to be predicted for any type of description element of the data to be encoded for the current block. Unlike the prior art, which, in order to select the signs to be predicted, jointly evaluates the prediction of a set of signs, the invention seeks to determine a subset of this set for which it is possible to predict each of the elements independently. values of others. To do this, it seeks to determine a sequence of the description elements of this subset that allows to achieve this. The invention is based on an entirely new and inventive approach that exploits the fact that the best prediction of a description element is the one knowing the values of all other elements and selects the subset of elements. description whose prediction individually maximizes a likelihood criterion. In this way, for the same number of predicted description elements as in the prior art, that is to say for the same processing complexity, the performances are improved.

Despite the fact that we do not know the other elements, we predict the identified element with the same probability of correct prediction as if we knew the true value of all the other elements of description. According to another aspect of the invention, the selection step comprises at least one iteration of a sequence of substeps implemented for a subset of the set of description elements of the current block and comprising:

Identifying a description element in the subset, such that the prediction value is independent of the values of the other description elements of the subset;

Constructing a sequence of description elements by concatenating the identified description element at the end of the sequence;

Updating the subset by deleting the identified description element;

The substep sequence being iterated for the updated subset, as long as the substep of identification has identified at least one description item.

Advantageously, it is iteratively constructed the ordered sequence of selected elements as long as possible from the initial subset. Following an iteration, the value or values of the identified sign (s) are set to their predicted value and used to identify other subset signs, knowing these values. If the final sequence has as many elements as the initial subset, it is because we have found a way to individually predict each element of the subset.

According to another aspect of the invention, the selection step further comprises a substep of defining an initial subset of the set of description elements, a description element being integrated into the subset according to a predetermined score associated with a coding context of the description element, said score being representative of a reliability level of the prediction of the description element, and in that the first iteration of the sub-elements previous steps is implemented for the initial subset.

One advantage is to take as a starting point to construct the ordered sequence, a subset of descriptive elements that are associated with a level of high prediction reliability, which guarantees a high compression ratio for the entropic coding of the prediction residuals of the selected description elements.

According to another aspect of the invention, when following the last iteration of the substep sequence, the constructed sequence does not include all the elements of the initial subset, the selection step further comprises a sub-step of restricting the subset by deleting at least one description element, depending on a predetermined score associated with a coding context of the description element, said score being representative of a level predictability of the description element, and that the next iteration of the substep sequence is applied to the restricted subset.

The invention seeks to integrate all the description elements of the subset in the sequence of description elements. If this is not possible with the initial subset, it is restricted, for example, by keeping only the elements associated with the highest scores, until a sequence comprising all the elements of the subset can be built.

According to another aspect of the invention, the substep of identifying at least one description element calculates the predicted value for the at least one description element and stores it in memory and in that the prediction step includes reading the stored value.

One advantage is that the prediction step does not need to recalculate this value. The computational load is optimized.

The method which has just been described in its different embodiments is advantageously implemented by a coding device of a digital image according to the invention, comprising the following units, which can be implemented for a current block: Processing of the current block intended to provide a set of data description elements to be encoded for the processed block;

- Prediction of the description elements of the selected subset;

Coding of the values of the unpredicted description elements;

Calculating a prediction indicator of the selected description elements from the predictions of the selected description elements and their original values, the indicator being intended to take a value in a group comprising:

a first value representative of a correct prediction;

a second value representative of an incorrect prediction;

Coding of the indicator values obtained for the predicted description elements;

According to the invention, the coding device is particular in that the selection unit is able to search, for a subset of the set of description elements, of a search order of the description elements of the sub-set. set for which a current element is predicted with the same value for all possible values of the remaining subset elements to be scanned and, when a browse order has been found, to select the subset.

Correlatively, the invention also relates to a method of decoding a digital image from a bit stream, said image being divided into a plurality of blocks processed in a defined order, the bit stream comprising coded data representative of elements description of the blocks of the image, said method comprising the following steps, implemented for a block, called current block:

Selecting a subset of description elements of the current block to be predicted from among a set of description elements of the block;

Decoding the non-selected description elements of the current block from coded data extracted from the bitstream;

Prediction of the values of the description elements of the selected subset; Decoding prediction indicator values of the selected description elements from coded data extracted from the bit stream, the indicator being adapted to take a value in a group comprising:

a first value representative of a correct prediction;

a second value representative of an incorrect prediction;

Calculating the decoded values of the description elements of the selected subset from the predicted values and the decoded prediction indicator values;

Reconstruction of the current block from the decoded values of all the description elements;

According to the invention, said method is particular in that, for the selected subset, the selection step searches, for a subset of the set of description elements, a search order of the description elements. of the subset for which a current element is predicted with the same value for all possible values of the remaining subset elements to be scanned and selects the subset, when a browse order has been found. Advantageously, the decoding method according to the invention exploits the coded data extracted from the bitstream to reproduce the step of selecting the description elements to be predicted in the current block, implemented by the coding method. the invention, the selection step comprises at least one iteration of a sequence of substeps, implemented for a subset of the set of description elements of the current block and comprising, for an iteration:

- Identification of a description element in the subset, such that the prediction value is independent of the values of the other description elements of the subset;

Updating the subset by deleting the identified description element; The sequence of substeps is iterated for the updated subset, as long as the substep of identifying the current iteration has identified at least one description element. According to yet another aspect of the invention, said step of selecting a subset of description elements comprises a sub-step of defining an initial subset of the set of description elements, a a description element being integrated into the initial subset based on a predetermined score associated with a coding context of the description element, said score being representative of a reliability level of the prediction of the description element, and in that the first iteration of the substeps is implemented for the initial subset.

According to yet another aspect of the invention, when following the last iteration of the substep sequence, the constructed sequence does not include all the elements of the initial subset, the selection step further comprises a sub-step of restricting the initial subset by deleting at least one description element, depending on a predetermined score associated with a coding context of the description element, said score being representative of a reliability level of the description element prediction, a decoding sub-step of the deleted description element using coded data extracted from the bitstream, and that the iterative sequence of substeps is applied to the initial restricted subset.

According to yet another aspect of the invention, a data description element of the block belongs to a group comprising:

- A prediction mode of the current block;

A mode of prediction of the motion vector estimated for the current block;

A coding mode of the current block;

A value of a coefficient of a residual of the current block;

Information representative of a significance of the value of the coefficient; and a sign of the coefficient.

Advantageously, the selected subset may comprise description elements of various types. The method which has just been described in its different embodiments is advantageously implemented by a device for decoding a digital image from a bit stream comprising coded data representative of said image, said image being divided into a plurality of blocks processed in a defined order according to the invention.

Such a device comprises the following units, able to be implemented for a block, called current block:

Prediction of the values of the description elements of the selected subset;

Decoding prediction indicator (IP) values of the selected description elements from coded data extracted from the bit stream, the indicator being adapted to take a value in a group comprising:

a first value representative of a correct prediction;

a second value representative of an incorrect prediction;

Reconstruction of the current block from the decoded values of all the description elements.

According to the invention, said device is particular in that, for the selected subset, the selection unit is able to search, for a subset of the set of description elements, of a travel order of description elements of the subset for which a current element is predicted with the same value for all possible values of the other elements of the subset that remain to be browsed and, when a browse order has been found, to select the sub-set -together.

Correlatively, the invention also relates to a signal carrying a bit stream comprising coded data representative of elements of description of blocks of pixels of a digital image, said pixel blocks being processed in a defined order.

The signal according to the invention is particular characterized in that said data encoded in the bit stream are obtained according to the coding method according to the invention. Correlatively, the invention also relates to a user terminal comprising a device for encoding a digital image and a device for decoding a digital image according to the invention.

The invention also relates to a computer program comprising instructions for implementing the steps of a method of encoding a digital image as described above, when this program is executed by a processor.

The invention also relates to a computer program comprising instructions for implementing the steps of a method of decoding a digital image as described above, when this program is executed by a processor.

These programs can use any programming language. They can be downloaded from a communication network and / or recorded on a computer-readable medium.

The invention finally relates to recording media, readable by a processor, integrated or not integrated with the coding device of a digital image and with the device for decoding a digital image according to the invention, possibly removable, respectively memorizing a computer program implementing an encoding method and a computer program implementing a decoding method, as described above.

6. List of figures

Other advantages and characteristics of the invention will emerge more clearly on reading the following description of a particular embodiment of the invention, FIG. 1 schematically illustrates a sequence of digital images to be encoded and the division into blocks of these images according to the prior art, as shown in FIG. FIG. 2 schematically shows the steps of a method of encoding a digital image according to a first embodiment of the invention; FIG. 3 details the processing step of a block implemented in the coding method according to the invention; FIG. 4 details the step of selecting a subset of description elements to be predicted, implemented by the coding method according to a second embodiment of the invention; Figure 5 schematically shows an embodiment of the selection step of the invention; Figure 6 schematically shows a decoded current block of a decoded digital image; FIG. 7 schematically shows the steps of a method of decoding a digital image according to a first embodiment of the invention; FIG. 8 details the step of selecting a subset of description elements to be predicted, implemented by the decoding method according to a second embodiment of the invention; and FIG. 9 shows an example of a simplified structure of a device for encoding a digital image and a device for decoding a digital image according to one embodiment of the invention.

7. Description of a particular embodiment of the invention

The general principle of the invention is based on the selection of a subset of data description elements to be encoded for a current block of a digital image, for which a sequence of the description elements can be defined according to wherein the prediction of the current description element is independent of the values of description elements that remain to be scanned. In relation to FIG. 1, an original video consisting of a sequence of M images II, 12,... IM, with M nonzero integer is considered. The images are encoded by an encoder, the encoded data is inserted a bit stream TB transmitted to a decoder via a communication network, or a compressed file FC, intended to be stored on a hard disk for example. The decoder extracts the coded data, then received and decoded by a decoder in a predefined order known from the encoder and the decoder, for example in the temporal order II, then 12, and then IM, this order being able to differ according to the mode of the decoder. production.

When encoding an image Im, with m integer between 1 and M, it is subdivided into blocks of maximum size which can in turn be subdivided into smaller blocks. Each block C will undergo an encoding or decoding operation consisting of a series of operations, including in a non-exhaustive manner a prediction, a residual calculation of the current block, a transformation of the pixels of the current block into coefficients, a quantization of the coefficients and entropy coding of the quantized coefficients. This sequence of operations will be detailed later.

The steps of the method of coding an image Im according to the invention will now be described in relation to FIG. 2.

In E0, we begin by selecting as current block C the first block to be processed. For example, this is the first block (in lexicographic order). This block has NxN pixels. During a step E1, a current block C is processed by implementing a coding scheme, for example as specified in the HEVC standard, in the document "ISO / IEC 23008-2: 2013 - High efficiency coding and media delivery in heterogeneous environments - Part 2: High efficiency video coding ", International Organization for Standardization, published November 2013. This processing step is intended to provide a set of ED elements for describing the data to be encoded for the current block C These description elements may be of various types, but not limited to: information relating to coding choices of the current block C, for example a mode of coding of the current block, such as the INTRA mode, INTER or SKIP, a mode of prediction of the current block, among the 35 modes of prediction of a block

INTRA, a prediction mode of an estimated motion vector for the block current, or the significance of an amplitude of a coefficient, known per se in HEVC; the data values to be encoded, such as the components of a motion vector, the amplitude or the sign of a coefficient; - etc

It is assumed that there are L cuts in possible blocks numbered from 1 to L, and that the cut used on the block C is the cut number I. For example, there may be 4 possible cuts, in blocks of size 4x4, 8x8 , 16x16, and 32x32.

In relation to Figure 6, the decoded current picture is designated ID. Note that in a video encoder, the ID image is (re) constructed in the encoder so that it can be used to predict the other pixels in the image sequence.

In relation with FIG. 3, an example of sub-steps implemented by this processing E1 of the current block C selected in accordance with the HEVC standard is detailed.

During a step El i, a prediction P of the original block C is determined. It is a prediction block constructed by known means, typically by motion compensation (block from a reference image previously decoded) in the case of a so-called INTER prediction, or INTRA prediction (block constructed from the decoded pixels immediately adjacent to the current block in the ID image). The prediction information related to P is encoded in the bit stream TB or compressed file FC. It is assumed here that there are K possible prediction modes m1, m2, ..., mK, with K nonzero integer, and that the prediction mode chosen for block C is the mode mk.

During a step E12, an original residue R is formed, by subtraction R = C-P from the prediction P of the current block C to the current block C.

During a step E13, the residue R is transformed into a transformed residue block, called RT, by a DCT transform or wavelet transform, both known to those skilled in the art and in particular implemented in the JPEG standards for the DCT and JPEG2000 for the wavelet transform.

In E14, the transformed residue RT is quantized by conventional quantization means, for example scalar or vector, into a quantized residue block RQ. This quantized block RQ contains NxN coefficients. As known in the state of the art, these coefficients are scanned in a predetermined order so as to constitute a vector one-dimensional RQ [i], where the index i varies from 0 to N ² -l. The index i is called the frequency of the coefficient RQ [i]. Classically, these coefficients are scanned in increasing frequency order, for example along a zigzag path, which is known from the JPEG fixed image coding standard. During a step E15, the amplitude information of the coefficients of the residual block RQ is encoded by entropy coding, for example by a Huffman coding or arithmetic coding technique. By amplitude is meant here the absolute value of the coefficient. Amplitude coding means are for example described in the HEVC standard and in the article by Sole et al, entitled "Transform Coefficient Coding in HEVC", published in the journal IEEE Transactions on Circuits and Systems for Video Technology, Volume 22, Issue: 12, pp. 1765 - 1777, December 2012. Classically, we can code for each coefficient representative information because the coefficient is non-zero. Then, for each nonzero coefficient, one or more information relating to the amplitude is encoded. CA amplitudes are obtained. At the end of step E1, the current block C has a set of data description elements ED to be encoded, among which are the quantized transformed residual coefficients RQ [i].

In relation with FIG. 2, during a step E2, a subset SE of this set, comprising the description elements to be predicted PED for the block C, is selected.

According to the invention, this subset is defined in a particular way, since it comprises only description elements which, when they are processed according to a particular travel order, can be predicted independently of the real values of the others. description elements of the set.

In relation with FIG. 4, the selection step E2 of a subset SE of description elements to be predicted according to one embodiment of the invention is now described.

In this example, description elements of a particular type are considered, for example signs of transformed and quantized coefficients of the current block RQ. Of course, the invention is not limited to this type of element and applies to any description element of the current block. During a first substep E21 we define an initial subset SEI description elements to predict.

Advantageously, one exploits the knowledge of a context Cxj associated with each coefficient among a plurality J of predetermined contexts, with J nonzero integer and j integer between 1 and J. Such a context is defined by at least one characteristic of coding of the coefficient or block from which it is derived.

Advantageously, the following characteristics are considered: the size of the quantized residue block RQ,

the amplitude of the quantized coefficient RQ [i],

the frequency of the coefficient or index i in the block RQ,

the mode of prediction of the current block mk among the K possible modes.

Indeed, the prediction of the sign is all the more reliable as the amplitude is high. Similarly, it was found that when the block is larger in size, the frequency of the coefficient lower, the prediction is more reliable. Finally, it has been found that the prediction is more reliable when the current block is associated with intra prediction of a certain type.

Alternatively, other contexts are conceivable. Thus, it is possible to take into account the type of image in which the current block is located, for example of the Intra or Inter type, known from the HEVC standard, as a function of the energy of the predictor P, or depending on the total number of non-zero coefficients in the current block.

During step E21, the signs of the coefficients of the block RQ to be selected, according to a predetermined score Sj, for the context Cxj associated with the coefficient RQ [i] considered, are selected.

Such a score Sj is representative of a reliability level of the prediction of the sign of the coefficient RQ [i].

For example, the score Sj takes values in a predetermined set, for example from 0 to 10. Alternatively, the score is a simple binary indication, one of which indicates that the sign can be predicted, and the other that the sign can not be predicted.

According to another variant, the scores Sj correspond to probabilities known a priori, dependent on the context Cxj associated with the coefficient RQ [i]. We have, in the encoder, a set of probabilities of correct detection of the signs of the coefficients RQ. For example, this set of probabilities is stored in memory.

These probabilities were constructed before encoding and decoding, either by statistical accumulation on a set of signals representative of the signals to be coded, or by mathematical calculation from hypotheses on the distribution of the signs of the coefficients. For a coefficient RQ [i] associated with the context Cxj, we can thus obtain the score Sj [i] by calculating the probability p [l] [mk] [i] [| RQ [i] | ] correct prediction of the sign of the coefficient RQ [i].

Advantageously, the signs to be predicted are selected by thresholding the scores with which they are associated. Thus, for each coefficient RQ [i] which has a sign (that is to say, which is not zero) and which is associated with a context Cxj of score Sj, the sign is predicted if and only if Sj > Th, where Th is a predetermined threshold, for example equal to 0.7. For example, the threshold Th is known from the encoder and the decoder.

According to one variant, the threshold Th may be chosen during encoding and written in the compressed file or in the bit stream comprising the coded data representative of the digital image Im. For example, if the unit that performs the encoding does not have enough computing resources at a given time, it is possible to increase this threshold Th so as to predict fewer signs, and therefore implement fewer calculations. It would also be possible to vary the threshold Th according to the content of the images to be encoded: an image containing a lot of content, such as large luminosity variations or large movements would use a high threshold, and an image containing little content such as that low luminosity variations or few movements, would use a lower threshold Th, so as to smooth the complexity or the memory necessary for the coding of each image. In relation to FIG. 5, a diagram is presented showing probability values of correct detection of description elements of a particular type, for example 10 signs of quantized coefficients RQ [i]. An EED set of description elements comprising ten elements of this type is considered. They are initially ordered according to the sequence order of the coefficients of the block implemented during the processing step El previously described. Th Threshold is used to select the initial SEI subset. In the example of FIG. 5, only the 4 index elements 1, 4, 5 and 8, whose correct prediction probability is greater than Th, are selected to form SEI. It is understood that the description elements of FIG. EED set that do not belong to the initial SEI subset, will not be predicted. They form a first NPED1 subset of unpredicted elements.

During a substep E22, it is sought if there exists, in SEI, a description element ED whose prediction is always the same, regardless of the value assigned to the other description elements of SEI. Concretely, this step consists in producing all the possible combinatorics for these other elements, then in observing what value of current Tc element is produced to predict this element, for each of these combinatorics.

By way of example, consider an EED set comprising 10 coefficients of the quantized residual block RQ, among which 4 signs are to be predicted. For example, suppose the RQ block equals {+8, +7, 0, -6, -3,0,0,2, -1,0}. Suppose also that the signs of the initial subset SEI are those of the 1 ^st , 4 ^th , 5 ^th and 8 ^th coefficients (of amplitudes 8, 6, 3 and 2 respectively). According to the invention, the prediction of the sign of the first coefficient 8 is first determined by setting the other three signs of SEI to a first value, for example {+, +, +} ■

For example, we define the combinatorial RQV0:

RQV0 = {+8, +7, 0, +6, + 3.0.0, + 2, -l, 0} RQV1 = {-8, +7, 0, +6, +3,0,0, + 2, -l, 0}

The blocks RQV0 and RQV1 are then decoded with conventional dequantization and inverse transforming means, the predicted block P is added to them to produce two virtual decoded blocks BDVO and BDV1. The likelihood of these blocks is tested with a likelihood criterion. The combination of signs corresponding to the virtual decoded block that maximizes the likelihood criterion is retained.

Advantageously, the likelihood criterion used is the minimization of the quadratic error along the boundary between the virtual decoded block and the previously decoded pixels.

In relation to Figure 6, there is shown a decoded picture ID and DVs virtual decoded block of size NxN pixels of the image, where DVs (n, m) is the value of DVs pixel of the block located on the n ^'th row and m ^th column of the block. The broken line F represents the boundary between the decoded virtual block and the rest of the image (previously decoded). ID (k, l) is the value of the ID pixel located on the k ^th line and the i ^th column of the image, and (lin, col) are the coordinates of the DVs block (coordinates of the pixel at the top left of DVS) in the image ID. We consider a "Side Matching" operator SM (3, B)), with 3 current image and B current block, defined as follows:

N-1 N-1

MS (3, B) = ((B (0, a) - 3 (Zin - 1, col + to) f + ((B (a, O) - 3 (Zin + a, col - 1)) ²

a = 0 a = 0

In FIG. 6, this amounts to forming the sum (xl-yl) ² + (x2-y2) ² + (x3-y3) ² + (x4-y4) ² + (x5-y4) ² + (x6-y5) ) ² + (x7-y6) ² + (x8-y7) ² . The optimal virtual decoded block DVopt is determined which minimizes this measurement:

DV _opt = argmin SM _DVs (ID, DV _s ) where ID represents the reconstructed image after decoding.

Alternatively, the likelihood criterion used is the minimization of the error with the predictor P. This consists in selecting the virtual decoded block which minimizes the error with the predictor P. The virtual residue associated with the optimal virtual decoded block is thus identified.

Suppose it is RQV0. We then obtain as a prediction a sign + for the first coefficient, when the other three signs of SEI are at {+, +, +}. According to the invention then tests all other possible combinations of the three values and signs of coefficients each time the sign of the prediction is calculated + ¹ coefficient

It is a question of comparing the values of likelihood obtained for the combinations of the values of the coefficients of indices 4,5 and 8 knowing + for the coefficient of index 1 with those obtained respectively with the same values of the coefficients of index 4, 5 and 8 knowing - for the ^1st coefficient.

For example, for the new combination {+, +, -} of the other three signs of SEI, we compare the likelihoods of the combinations: RQV2 = {+8, +7, 0, +6, + 3,0,0, - 2, -1,0}

RQV3 = {-8, +7, 0, +6, + 3.0.0, -2, -1.0} and it is verified that the combination RQV2 is more likely than the combination RQV3.

If the likelihood is maximum for all possible combinations of the coefficients 4.5 and 8 while the sign of the ^1st coefficient is predicted with the + sign, then was identified by ¹ coefficient as predictable regardless of the values assigned to the other.

In other words, if the predicted value for the current element is always the same, then we consider in E23 that the current element is identified as the first element of a Seq sequence that is built in E24. In E23, it was previously verified that the sequence under construction had not already been tested.

In E25, the current subset SEI is updated by deleting the identified element from the initial subset SEI0. For the rest of the selection, the identified element is fixed at its predicted value.

The process is iterated by repeating steps E22 to E25 as long as the subset SE comprises more than one element and as long as it is possible to identify a new ED element whose prediction is always the same, regardless of the value assigned. other elements of SE description.

If the process stops because step E22 has not identified a new description element while the subset SE is not empty, the sequence Seq constructed in E24 is incomplete because it does not include all elements of the initial subset. This is the case for the first sequence Seq 1 obtained in the example of FIG. initial subset SEI0 with 4 elements. It includes only two elements of indices 1 and 8.

Advantageously, the process is restarted to try to construct another ordered sequence Seq from the subset SEI0, by identifying another element ED in the first position of the sequence. If the process ends, while this other sequence is not complete, then the process is restarted at E27, until all the possible sequences for the initial SEIO subset have been tested.

Seq is reset to E28.

If all the ordered sequences have been tested for a given SEI without obtaining any complete sequence, it is considered that the selection process failed, in the sense that it did not find any order of the ED elements in SEI allowing them to be processed. predict individually.

The initial SEIO subset is restricted during a step E29. Advantageously, the element ED which is associated with the lowest score Sj is deleted and a new initial subset SEI1 is obtained.

The process is repeated for the new SEI1 initial subassembly.

This is the case in the example of FIG. 5. The element of index 5 is removed and the selection process of a subset of elements to be predicted SEI1 reduced to 3 index elements is iterated. , 4 and 8.

If it is still not possible to obtain a complete sequence for the new SEI1, SEI1 is further restricted to SEI2. Note in this respect that when SEI is reduced to one element, the algorithm stops, since the prediction of this element is necessarily independent of the others.

If a complete Seq sequence has been found for an SEI subset, then all ED elements in this set can be predicted independently of each other in the order of the order of the elements of the found sequence. Advantageously, the predicted values are memorized for each of the elements of the Seq sequence. found. Thus, they can be reused during the actual prediction stage.

This is the case of the sequence Seq2 found in the example of FIG. 5. It comprises in the order the three elements ED, of indices 8.1 and 4. The other 7 elements of the set EED belong to the SENP subset of unpredicted elements. The original SEI0 description elements that have been removed from the final SE = SEI1 subset form a second subset of NPED2 elements that will not be predicted. In another embodiment, description elements are selected to predict of another type. In particular, the description element M indicative of the INTRA / INTER prediction mode is considered (in the HEVC standard, such a description element is called "pred_mode_flag"), the description element A indicative of the amplitude of the first quantized residual coefficient for the current block (in the HEVC standard, such a description element is called "coeff_abs_level_remaining") and the description element T indicates the use or not of an inverse transform (in the HEVC standard, such a description element is called "transform_skip_flag").

For the current block, the starting set consists of the description elements {M, A, T}.

It is considered in this example that, for the current block C, and depending on the contextual information, it is found in step E21 that the score of T is less than the necessary threshold Th, while M and A have a higher score. . The initial subset SEI = SEI0 is therefore {M, A}. In the course of step E22, SEI is searched for whether there is a description element whose prediction is independent of the values of the other syntax elements.

First, we examine whether the prediction of the amplitude A is different according to whether the prediction mode M is INTER or INTRA. Suppose we find that the amplitude is different in both cases. So, this means that a sequence Seq (each element of which has a prediction independent of the following elements of the series) starting with A does not exist in SEI. We will then test if it is possible to find such a sequence starting with M. We will therefore determine if the prediction mode INTER or INTRA is the same regardless of the value of the amplitude A. Suppose that for the block current considered, the predicted prediction mode is always the INTRA mode, regardless of the value of the amplitude A. We can then choose the description element M as the first element of the Seq series. In addition, there remains only one element of description in SEI, the element A. We know that we can also predict the value of A independently of the remaining description elements - since there are more. The sequence Seq is therefore composed of the elements M, A in this order. The selection step E2 which has just been described can advantageously be implemented in the form of a recursive function f (C, LP), where C denotes the current block to be coded and LP the n-tuple of the descriptive elements. of the initial subset SEI, with n integer between 1 and N, with N not zero. An example of a pseudo code of this function f is given in the appendix.

At the end of this step, we have therefore constructed an ordered sequence Seq of description elements to predict of length equal to that of the subset SE defined by the selection step E2.

During a step E3, the description elements of the first and second subsets NPED1 and NPED2 of the set EED which do not belong to the selected subset SE are conventionally encoded. This step implements coding techniques known to those skilled in the art. For the signs of the coefficients RQ [i] for example, it is known in particular of the HEVC standard, in particular of the article by Sole et al., Already mentioned, the fact of transmitting each sign in the form of a binary element 0 or 1, with a convention associating one with the plus sign and the other with the minus sign.

During a step E4, the ED elements of the subset SE obtained are predicted. This is done simply by calling the predicted value stored during the step E2 of selecting the signs to be predicted.

During a step E5, for each description element to be predicted PED, an information representative of the difference between the prediction of the description element and the real value of this element, called the IP prediction indicator or the residual of PED, is calculated. the PED description element.

Thus, taking the example of FIG. 5, considering that the elements ED are signs, the prediction of the elements of indices 8, 1, 4 of the sequence Seq2 of FIG. 5 is {-, -, +} while the true signs are {+, -, +}. By convention, the prediction indicator IP is set to 1 when the prediction is correct and to 0 when the prediction is incorrect, so we obtain {0, 1, 1}.

During a step E6, the values of the IP prediction indicator for each PED element to be predicted are coded by a known entropy coding technique, such as for example a Huffman coding, arithmetic coding or CABAC coding such as 'used in the HEVC standard. A CIP value of the coded prediction indicator is obtained.

According to the invention, since only the description elements which are associated with a representative score of a sufficient level of reliability are predicted, the prediction indicator more often takes the value 1 than the value 0. This is set to profit by entropic coding to reduce the size of the compressed signal.

Advantageously, the entropy coding takes into account the score Sj associated with the predicted sign for coding the IP indicator. For example, in the embodiment of the invention in which the score has a value between 0 (low reliability of the prediction) and 10 (high reliability of the prediction), the entropic coding of the indicators is parameterized taking into account the score, so as to exploit the more or less uniform distribution of the indicators. For example, CABAC-type entrapic coding, known from the HEVC standard, is used by initializing the probabilities used in CABAC based on the predetermined scores.

During a step E7, illustrated by FIGS. 2 and 3, the decoded block BD corresponding to the quantized residual block RQ is constructed by applying to it the dequantization and inverse transforming steps known to those skilled in the art. A decoded residue block RD is obtained. The predictor block P is added to RD to obtain the decoded block D.

During this step we also add the decoded block D to the reconstructed image ID. This makes it possible to have in the encoder a decoded version of the current image. This decoded version is used in particular during the step of constructing a prediction of the signs selected to be predicted. During a step E8, the coded data, that is to say the indicators of the predicted description elements PED and the unpredicted description elements NPED are inserted in the bit stream TB or in the compressed file FC.

At the end of the processing of the current block C, E9 is tested whether the current block C is the last block to be processed by the coding unit, taking into account the coding run order defined above. If yes, the coding unit has finished processing. If not, the next step is the step of selecting the next block E0. This block becomes the current block to be processed, and the next step is the step E1 of processing the current block, already described. In an exemplary embodiment of the invention, the context Cxj associated with a score Sj for a sign-type description element, depends on the size of blocks I (out of 4 possible sizes, as previously described), the prediction mode intra mk among 35 possible prediction modes (as described in the HEVC standard mentioned above), the frequency i (among 16, 64, 256 or 1024 possible frequencies, depending on the size of the blocks), and Tamplitude | RQ [i] | (which can take 256 possible values when encoded on 8 bits). In this example, the number J of contexts used is equal to 35x (16 + 64 + 256 + 1024) x256 = 12185600.

A preliminary examination on typical image sequences makes it possible to calculate a probability of correct detection of the sign for each of the contexts Cxj. This probability is the score Sj associated with each context Cxj, which makes it possible to select the signs to be predicted from a Th threshold of 0.7, as previously described. Thus, a compression gain of 1 to 2% is observed compared to the state of the art.

The TB bitstream is intended to be presented at the input of a decoder, local or remote. For example, a signal carrying the bit stream is transmitted to the decoder via a communication network.

In relation to FIG. 7, the steps of the method for decoding a coded digital image according to an exemplary embodiment of the invention are now presented. It is assumed that a bit stream TB has been received by a decoding device implementing the decoding method according to the invention. In DO, we begin by selecting as current block C the first block to be processed. For example, this is the first block (in lexicographic order). This block has NxN pixels.

During a step D1, the method reads in the bit stream TB the coded data relating to the description elements of the unpredicted current block belonging to the subassembly NPED1 and decodes them. At the end of this step, it knows the decoded values of the description elements of the block that have not been predicted because they were associated with a score Sj less than a predetermined Th threshold.

During a step D2, it implements the step of selecting the description elements to be predicted, already described in detail for the coding method in relation with FIGS. 3 and 4.

In relation to FIG. 8, a particular embodiment of this selection step is presented.

In D21, we define an initial subset SEI = SEI0, which initially contains all the description elements that have not been decoded in the previous step, then we search in this subset for a traversal order of the elements, according to which the prediction of the current element takes the same value for all the possible values of the elements of the subset that remain to be scanned. It can therefore be predicted independently of the value of others. In D22, it is sought to identify a first description element of the subset SEI, which can be predicted independently of the value of the others. If an element is identified (D23), its predicted value is stored in memory. This element is placed in the first position of a sequence Seq (D25), it is removed from the initial SEI subset and it is fixed at the predicted value. The substeps D22 to D25 are iterated for the subset SE obtained to identify a second element, then a third, etc., for the sequence Seq, as long as the subset SE has more than one element.

If at the end of this loop, the sequence Seq includes all the elements of the subset, the selection step D22 was successful. All elements of the subset SEI = SEI0 will be predicted in D3 according to the order found.

Otherwise, the subset SEI is restricted to D29, deleting a description element, for example that which is associated with the lowest score. If the new subset SEI obtained comprises at least two elements (test in D210), it serves as a basis for a new iteration of substeps D22 to D211 which have just been described in order to construct an ordered sequence for this new subset.

The description element deleted in D29 was therefore not predicted by the coder. It belongs to a second subset NPED2 of unpredicted description elements. It is therefore decoded, in D211, from coded data extracted from the TB bitstream. Indeed, its decoded value will be exploited by the next iteration of the identification sub-step D22 to predict the current description element knowing the values of all the other description elements of the current block.

The process is repeated until a complete Seq sequence has been found for an initial SEI subassembly or the restricted SEI subset contains more than one element. Indeed, only one element may need to be predicted independently of the others.

At the end of step D2, an ordered sequence of PED description elements to be predicted in a subset SEI comprising at least one element has therefore been determined. The original UTE description elements that have been removed from the final UTE form a second subset of NPED2 elements that will not be predicted.

Taking the example of FIG. 5, the Seq2 sequence found for the subset SEI comprises in the order the three elements ED, with indices 8.1 and 4. The first set NPED1 comprises the 6 index elements 2 , 3, 6, 7, 9 and 10. The second subset NPED2 includes the element of index 5.

In D3, the PED description elements of the sequence Seq found are predicted. Advantageously, the method will simply seek in memory the values predicted during the previous step D3. In D4, the method reads in the bit stream TB the coded data relating to the prediction residuals of the predicted values for the description elements of the Seq sequence. As previously described for the coding method, this residue is an information representative of the difference between the description element prediction and the real value of this element, called the IP prediction indicator. The CIP coded data read in the bitstream corresponds to the coded values of this prediction indicator and must first be the subject of an entropy decoding corresponding to the inverse operation of that performed by the coding method. Thus, if we take again the example of Figure 5, considering that the elements ED are signs, the decoded data read in the bitstream are {0, 1, 1}, which indicates that the prediction of the first sign of the sequence is incorrect and those of the two following signs are correct. In D5, the method calculates the actual values of the predicted description elements

PED from their predictions and decoded IP indicators.

In the example of FIG. 5, the prediction of the index elements 8, 1, 4 of the sequence Seq2 of FIG. 5 is {-, -, +} is corrected by the decoded IP indicators {incorrect, correct, correct} to obtain the true signs {+, -, +} In D6, the calculated values of the predicted and unpredicted description elements are used to reconstruct the current block C. Notably, this step implements techniques known at night. :

dequantization of the current block RQ 'to obtain a dequantized block. This is done by means adapted to the quantization used during the coding (scalar dequantization, vector dequantization ...);

Inverse transformation of the dequantized current block RT '; and

Construction of the decoded block by adding the residue R 'to the prediction P of the current block C.

The reconstructed block C is integrated with the image being decoded.

During a step D7, it comes to test whether the current block is the last block to process the decoder, given the order of travel defined above. If so, the decoding process has finished processing. If no, the next step is the step of selecting the next block D0 and the decoding steps D1 to D7 previously described are repeated for the next block selected.

It will be noted that the invention which has just been described can be implemented by means of software and / or hardware components. With this in mind, the terms "module" and "unit" used in this document may correspond to either a software component, a hardware component, or a set of hardware and / or software components, able to implement the function (s) described for the module or unit concerned.

With reference to FIG. 9, an example of a simplified structure of a device 100 for encoding a digital image and a device 200 for decoding a bit stream according to the invention is now presented. The device 100 implements the coding method according to the invention which has just been described in relation with FIG. 2.

For example, the device 100 comprises a processing unit 110, equipped with a processor μ ΐ, and driven by a computer program Pg l 120, stored in a memory 130 and implementing the method according to the invention. At initialization, the code instructions of the computer program Pgi 120 are for example loaded into a RAM memory before being executed by the processor of the processing unit 110. The processor of the processing unit 110 sets implement the steps of the method described above, according to the instructions of the computer program 120. In this exemplary embodiment of the invention, the device 100 comprises at least one processing unit PROC of a current block intended to provide a set description elements of the current block, a SEL unit for selecting a subset of elements to be predicted from the set provided for the current block, an NPED COD unit for encoding unselected elements, a PRED unit of elements of the selected subset, a CALC unit for calculating a correct prediction indicator of a description element and a COD unit for encoding the calculated indicator values for the elements predicted ents.

The device 100 further comprises a unit M 1 for storing the coding contexts of the coefficients, predetermined scores associated with each of these contexts, and predicted values for the selected description elements.

These units are driven by the μΐ processor of the processing unit 110.

Advantageously, such a device 100 can be integrated with a user terminal TU. The device 100 is then arranged to cooperate at least with the following module of the terminal TU: a data transmission / reception module E / R, through which the bit stream TB or the compressed file FC is transmitted in a telecommunications network, for example a wired network or a radio network.

The decoding device 200 implements the decoding method according to the invention which has just been described in relation to FIG. 7.

For example, the device 200 comprises a processing unit 210, equipped with a processor μ2, and driven by a computer program Pg2 220, stored in a memory 230 and implementing the method according to the invention.

At initialization, the code Pg ₂₂₂₀ computer program instructions are loaded for example in a RAM memory before being executed by the processor of the processing unit 210. The processor of the processing unit 210 implements the steps of the method described above, according to the instructions of the computer program 220.

In this exemplary embodiment of the invention, the device 200 comprises at least one DEC NPED unit for decoding the unpredicted description elements of a current block C from the coded data extracted from the bit stream, a selection unit SEL '. of a subset of description elements to be predicted in the set provided, a prediction unit PRED 'for predicting the values of the description elements of the selected subset, a DEC decoding unit of prediction indicator values description elements selected from coded data extracted from the bit stream, a CALC PED unit for calculating the decoded values of the description elements of the subset selected from the decoded prediction indicator values and their predictions and a unit RECONST of reconstruction of the current block C. The device 200 further comprises a unit M2 for storing the coding contexts of the coefficients, predetermined scores associated with each of these contexts, and predicted values for the PED description elements selected for a block C. These units are driven by the processor μ2 of the processing unit 210. Advantageously, such a device 200 may be integrated with a user terminal TU. The device 200 is then arranged to cooperate at least with the following module of the terminal TU: a data transmission / reception module E / R, through which the bit stream TB or the compressed file FC is received from the network. telecommunications.

It goes without saying that the embodiments which have been described above have been given for purely indicative and non-limiting purposes, and that many modifications can easily be made by those skilled in the art without departing from the scope. of the invention.

ANNEX

recursive function f (C, LP)

{

if LP is empty, return true.

if not

for each EDk element of LP

{

is the prediction of EDK constant regardless of the value of the other ED elements of LP?

If not, return false

if yes

{

Remove the spk syntax element from LP

We memorize the prediction of spk.

In block C, the syntax element spk is set to its real value.

Return f (C, LP)

}

beginning

Lfinal list = LP

As long as Lfinal is not empty

{

if f (Lfinal) end;

if not

{

Remove from Lfinal the element that has the lowest probabilty p (s) of correct prediction

}

End

Claims

A method of coding a digital image, said image (Im) being divided into a plurality of blocks of pixels (C) processed in a defined order, said method comprising the following steps, implemented for a current block:

- Processing (El) of the current block for providing a set of description elements of the processed block;

Selecting (E2) a subset of description elements to be predicted from the provided set;

Prediction (E4) of the description elements of the selected subset;

Coding (E3) of the values of the unpredicted description elements;

Calculating (E5) a prediction indicator (IP) of the selected description elements from the predictions of the selected description elements and their original values, the indicator being intended to take a value in a group comprising:

a first value representative of a correct prediction;

a second value representative of an incorrect prediction;

Coding (E6) of the indicator values obtained for the predicted description elements;

said method being characterized in that the selecting step (E2) searches, for a subset of the set of description elements, a browse order of the subset description elements for which a prediction of the subset current element takes the same value for all the possible values of the other elements of the subset that remain to be browsed and selects the subset, when a browse order has been found.

2. A method of coding a digital image according to claim 1, characterized in that the selection step comprises at least one iteration of a sequence of substeps comprising, for an iteration, implemented for a sub-step. set of description elements of the current block: Identifying (E22) a description element in the subset, such that the prediction value is independent of the values of the other description elements of the subset;

Constructing (E24) a sequence of description elements by concatenating the identified description element at the end of the sequence;

Update (E25) of the subset by deleting the identified description element;

A method of coding a digital image according to claim 2, characterized in that the selecting step (E2) further comprises a sub-step (E21) for defining an initial subset (SEI) of the set of description elements, a description element being integrated in the subset according to a predetermined score (Sj) associated with a coding context (Cxj) of the description element, said score being representative of a reliability level of the prediction of the description element, and in that the first iteration of the preceding substeps is implemented for the initial subset.

A method of coding a digital image according to one of claims 2 and 3, characterized in that, when following the last iteration of the substep sequence, the constructed sequence does not include all the elements of the initial subset, the selection step (E2) further comprises a sub-step (E29) for restricting the subset by deleting at least one description element, depending on a predetermined predetermined score to a coding context of the description element, said score being representative of a reliability level of the prediction of the description element, and in that the next iteration of the sequence of substeps is applied to the restricted subset.

Method for coding a digital image, according to one of Claims 2 to 4, characterized in that the sub-step of identifying at least one element of description computes the predicted value for the at least one description element and stores it in memory and in that the prediction step (E4) comprises reading the stored value.

A device (100) for encoding a digital image, said image (Im) being divided into a plurality of blocks of pixels processed in a defined order, said device comprising the following units, which can be implemented for a current block ( VS) :

Processing (PROC) of the current block for providing a set (EED) of description elements of the processed block;

Selection (SEL) of a subset (SE) of description elements to be predicted from the provided set;

Prediction (PRED) of the description elements of the selected subset; Coding (COD NPED) of the values of the unpredicted description elements; Calculation (IP CALC) of a prediction indicator (IP) of the selected description elements from the predictions of the selected description elements and their original values, the indicator being intended to take a value in a group comprising:

a first value representative of a correct prediction;

a second value representative of an incorrect prediction;

Coding (COD IP) of the indicator values obtained for the predicted elements of description;

said device being characterized in that the selection unit is able to search, for a subset of the set of description elements, a travel order of the description elements of the subset for which a prediction of a current element takes the same value for all the possible values of the other elements of the subset that remain to be scanned and, when a browse order has been found, to select the subset.

Method for decoding a digital image (Im) from a bit stream (TB), said image being divided into a plurality of processed blocks in a defined order, the bit stream including coded data representative of picture block description elements, said method comprising the following steps implemented for a block (C ), said current block:

Selection (D2) of a subset (SE) of description elements of the current block to be predicted (PED) from among a set of description elements of the block; Decoding (D1) of the unselected description elements of the current block from encoded data extracted from the bit stream;

Prediction (D3) of the values of the description elements of the selected subset;

Decoding (D4) prediction indicator (IP) values of the selected description elements from encoded data extracted from the bit stream, the flag being for taking a value in a group comprising:

at. a first value representative of a correct prediction;

b. a second value representative of an incorrect prediction;

Calculating (D5) the decoded values of the description elements of the subset selected from the decoded prediction indicator values;

Reconstruction (D6) of the current block from the decoded values of the set of description elements; said method being characterized in that the selection step (D2) searches, for a subset of the set of description elements, a browse order of the description elements of the subset for which a prediction of a current element takes the same value for all the possible values of the other elements of the subset that remain to be browsed and selects the subset, when a browse order has been found.

Method for decoding a digital image according to claim 7, characterized in that the selection step (D2) comprises at least one iteration of a sequence of substeps comprising, for an iteration: Identifying (D22) a description element in the subset, such that the prediction value is independent of the values of the other description elements of the subset;

Constructing (D24) a sequence (Seq) of description elements by concatenating the description element identified at the end of the sequence;

Updating (D25) the subset by deleting the identified description element;

The sequence of substeps is iterated for the updated subset, as long as the identification substep implemented during the current iteration has identified at least one description element.

A method of decoding a digital image (Im) from a bit stream (TB) comprising encoded data representative of said image according to claim 8, characterized in that said step (D2) of selecting a sub-frame set of description elements includes a subset (D21) of initial subset definition (SEI) of the set of description elements, a description element being integrated with the initial subset, based on a predetermined score associated with a coding context of the description element, said score being representative of a reliability level of the prediction of the description element, and in that the first iteration of the substeps is implementation for the initial subset (SEI).

10. The method of decoding a digital image according to one of claims 8 or 9, characterized in that, after the last iteration of the sequence of substeps, the sequence (Seq) constructed does not include not all the elements of the initial subset, the selection step (D2) further comprises a sub-step (D29) for restricting the initial subset by deleting at least one description element, depending on a predetermined score associated with a coding context of the description element, said score being representative of a reliability level of the prediction of the description element, a substep (D211) for decoding the description element, deleted description element using extracted coded data of the bit stream and that the next iteration of the substep sequence is applied to the restricted initial subset.

11. The method of decoding a digital image according to one of claims 7 to 9, characterized in that a data description element of the block belongs to a group comprising at least:

- A prediction mode of the current block;

A mode of prediction of the motion vector estimated for the current block;

- A coding mode of the current block;

A value of a coefficient of a residual of the current block;

- information representative of a significance of the value of the coefficient; and

- A sign of the coefficient.

A device (200) for decoding a digital image (ID) from a bit stream (TB), said image being divided into a plurality of processed blocks in a defined order, the bit stream comprising representative encoded data elements of description of the blocks of the image, said device comprising the following units, which can be implemented for a block (C), called current block:

Selection (SEL ') of a subset (SE) of description elements of the current block to be predicted (PED) from among a set of description elements of the block; Decoding (DEC NPED) the non-selected description elements of the current block from coded data extracted from the bit stream;

Prediction (PRED ') of the values of the description elements of the selected subset;

Decoding (DEC IP) of prediction indicator (IP) values of the selected description elements from encoded data extracted from the bit stream, the flag being for taking a value in a group comprising:

vs. a first value representative of a correct prediction;

d. a second value representative of an incorrect prediction;

Calculating (CALC ED ') the decoded values of the description elements of the selected subset from the predicted values of the description elements and the decoded prediction indicator values; Reconstruction (RECONST) of the current block from the decoded values of all the description elements; said device being characterized in that the selection unit (SEL ') is able to search, for a subset of the set of description elements, a running order of the description elements of the subset for which a prediction of a current element takes the same value for all the possible values of the other elements of the subset that remain to be scanned and, when a browse order has been found, to select the subset.

A signal carrying a bit stream (TB) comprising encoded data representative of pixel block description elements of a digital image, said pixel blocks being processed in a defined order, characterized in that said encoded data is obtained according to the coding method according to one of claims 1 to 5.

14. User terminal (TU) characterized in that it comprises a coding device of a digital image according to claim 6 and a device for decoding a digital image according to claim 12.

15. Computer program comprising instructions for implementing the method of coding a digital image according to one of claims 1 to 5, when executed by a processor.

16. A computer program comprising instructions for implementing the method of decoding a digital image according to one of claims 7 to 11, when executed by a processor.