WO2004047452A1

WO2004047452A1 - Method and system for measuring degradations of a video image introduced by digital broadcast systems

Info

Publication number: WO2004047452A1
Application number: PCT/FR2003/003133
Authority: WO
Inventors: Nathalie Montard
Original assignee: Telediffusion De France
Priority date: 2002-11-15
Filing date: 2003-10-22
Publication date: 2004-06-03
Also published as: FR2847412B1; AU2003285436A1; FR2847412A1

Abstract

The invention concerns a method for measuring degradations of a digital video image introduced by an image transmission system by macro blocks assembling several image blocks, consisting in: performing a pre-filtering (12) by applying to each transmitted image macro block a statistical test to determine whether there is equality of property between the macro block and the block adjacent the macro block, and in retaining the macro blocks whereof the property is not equal to that of its adjacent blocks; for each macro block retained, determining (13) whether the blocks of the macro block have a property, based on several direction criteria, similar to that of the neighbouring blocks, and judging the macro block to be erroneous if it is determined that it is not similar according to several of the direction criteria, and determining (14) an index of degradation of the image based on the number of macro blocks considered to be erroneous in the image.

Description

METHOD AND SYSTEM FOR MEASURING DEGRADATIONS OF A VIDEO IMAGE INTRODUCED BY DIGITAL BROADCASTING SYSTEMS.

The present invention relates to a method and a system for measuring the degradations of a video image, introduced by digital broadcasting systems.

It applies in particular, but not exclusively, to the field of broadcasting networks for low-speed or very low-speed digital audiovisual signals. It applies in particular to monitoring the quality of service of a network for broadcasting digital audiovisual signals.

The digitization of video signals makes it possible to copy, store and transmit this type of information while maintaining a constant image quality. However, the large amount of information conveyed by video images in practice requires the use of digital compression methods to reduce the bit rate.

A widely used compression method in the video field is described in ISO / IEC 13918 MPEG2. This algorithm is said to be of the "lossy" type because the image restored after decoding is not identical to the original. Such an algorithm is based on a division of the image into blocks and on the application of a transform, for example of the type transformed into a discrete cosine to the pixels of each block, which makes it possible to obtain a frequency representation of the amplitude the luminance of the pixels in the form of as many coefficients as pixels in the block. According to the MPEG2 standard, the coded images are transmitted in the form of a bit stream in which the blocks are grouped in macro blocks of 2 x 2 blocks, so that if a transmission error appears, it generally affects at least an entire macro block.

In order to maintain an acceptable quality for the end viewer, the compression algorithms take into account the perception properties of the human vision system. However, the bit rate constraints imposed by the transmission systems require the application of compression rates which influence the quality of the image perceived by the viewer. It turns out that the extent of the degradations liable to be undergone by a coded image or a sequence of such images depends both on the compression rate used during the coding of the images and on the complexity thereof, in particular in which particularly concerns the movement of objects, brightness, and texture.

Among the degradations which appear in the images coded by the MPEG technique which are transmitted at low speed for the broadcasting of digital television or at lower speed for other multimedia applications, one can cite among the most perceptible the effects of (macro ) misplaced or wrong blocks, loss of information, appearance of "exotic" contours, etc. The effect of a disturbance depends on the level of relevance of the data it affects or on the data structure used. It can thus affect the synchronization words or the motion vector. In the case where the images are coded with prediction, a disturbance may affect the base images for the predictions. Small transmission errors can result in more or less localized effects on the image, in particular in the form of erroneous or incorrectly positioned blocks or macro blocks in the image. In the event of a major disturbance, they can cause difficulties in accessing information, and in particular, service interruptions for a more or less long period, which can lead to image crashes or freezes.

It therefore appears necessary to permanently assess the quality of the images broadcast. There are widely used subjective assessment methods for this purpose, using human appreciation. These methods are however cumbersome to implement and cannot be used in real time on a functioning network.

There are also so-called "with reference" methods based on the comparison of the image whose quality is to be evaluated with a reference image. The reference image is generally that which corresponds to the image to be analyzed before its coding and / or its transmission. This solution proves to be impractical because it requires access to one or more reference images, and therefore the transport of these reference images to the place of reception of the image to be analyzed.

Other solutions called "without reference" allow to automatically analyze images without having to make a comparison with reference images. The efficiency and robustness of each of these solutions lies in their ability to discriminate between the objects in a scene and the damage that may be found there.

Thus, the publication "Fuzzy similarity measures for the detection of image transmission errors with vector quantization" by B. Solaiman, R. Pindiah, O. Aitsab, G. Cazuguel and C. Roux, Dept. Image and Information Processing, Telecom Bretagne, CORESA, March 1997, proposes to perform a fuzzy similarity measure for the detection of errors in images coded by vector quantization. To this end, it uses a dictionary of code words or reference sub-images, correlated and ordered so as to assign similar similarity indices to similar blocks. The image to be transmitted is made up of blocks of sub-images and the index of the nearest word is transmitted instead of the entire block. The received images are degraded when transmission errors occur in the transmitted indices. For the current block, we consider all eight blocks which surround it, characterized by their luminance average. A measure of overall similarity of the central block with respect to its neighborhood is evaluated to determine whether the central block is likely to be erroneous.

In general, the methods without reference of the prior art use filters only highlighting a certain type of image content: borders or high frequencies, or representing only part of the content of the image. 'picture. Consequently, the degradation measurements thus carried out are of limited relevance in terms of identification or discrimination of the degradations. In particular, these methods often poorly detect degradations among the objects of a scene. As a result, they have a high error rate.

The present invention aims to eliminate these drawbacks. This objective is achieved by providing a method for measuring degradations of a digital video image introduced by a transmission system, the image being coded using coding processing involving a decomposition of the block image and a block transform calculation, the image being transmitted by macro blocks grouping together several blocks. According to the invention, this method comprises steps consisting in:

- perform a pre-filtering consisting in applying to each macro block of the image transmitted a statistical test to determine if there is equality between a property of the macro block and a corresponding property of the blocks adjacent to the macro block, and to retain the macro blocks whose property is not equal to that of its blocks adjacent, - for each macro block selected, determine whether the blocks of the macro block have a similar property, according to several direction criteria, to that of neighboring blocks, and consider that the macro block is erroneous if it is determined not similar according to several direction criteria, and - determining an image degradation index as a function of the number of macro blocks considered erroneous in the image.

According to a feature of the invention, the direction criteria comprise a first criterion, consisting in comparing a property of each macro block retained with that of neighboring blocks having a common border with the macro block, a second criterion consisting in comparing a property of each macro block retained with that of neighboring blocks located on diagonals of the macro block, and a third criterion consisting in comparing a property of each macro block retained with that of neighboring blocks delimiting the angles of the macro block, a macro block being considered erroneous s 'it is determined not similar according to at least two of the three criteria.

According to another feature of the invention, this method comprises a prior synchronization step consisting in determining the coding grid of the coded image, in order to find the decomposition into coding blocks of the image, used during the coding of the image. 'picture.

According to yet another feature of the invention, determining the similarity of the property of a macro block of the image with that of neighboring blocks consists in determining a fuzzy similarity index and in comparing the similarity index determined with a threshold depending on the value of the pixels of the image in an area in which the macro block is located.

Advantageously, the fuzzy similarity index is obtained using a membership function applied to an average luminance difference between the macro block and neighboring blocks of the macro block.

Preferably, the membership function is defined by the following formula: μ _c iose ( ^Δ ) = ι - ^ l ^Δ l

Δ being an average luminance difference between the macro block and neighboring blocks of the macro block, and M being the maximum admissible value of average luminance.

According to yet another feature of the invention, a macro block is considered to be erroneous according to the first criterion if it has at least two boundaries along which the blocks of the macro block have an average luminance sufficiently different from that of adjacent blocks in the vicinity. of the macro block.

According to yet another feature of the invention, a macro block is considered to be erroneous according to the second criterion if at least three of the corner blocks have an average luminance sufficiently different from that of respective diagonal blocks in the vicinity of the macro block.

According to yet another feature of the invention, a macro block is considered to be erroneous according to the third criterion if at least two corner blocks of the macro block have an average luminance sufficiently different from that of blocks in the vicinity of the macro block, adjacent to the respective angle.

Advantageously, the pre-filtering consists in applying to each macro block of the image a statistical equality test to determine whether there is equality of mean luminance between the macro block and the neighboring blocks of the macro block.

The invention also relates to a system for measuring degradations of a digital video image introduced by a transmission system, comprising means for implementing the method defined above.

A preferred embodiment of the invention will be described below, by way of nonlimiting example, with reference to the appended drawings in which:

FIG. 1 schematically represents the measurement system according to the invention, integrated in an image processing chain;

Figure 2 shows in more detail the measurement system shown in Figure 1; Figure 3 shows in more detail a part of the system shown in Figure 2;

FIG. 4 shows an image part composed of 16 blocks to illustrate the method according to the invention.

FIG. 1 represents a system 1 for measuring degradations according to the invention, which is intended to be used as a probe capable of being applied at any point of a chain for receiving digital images, such as a chain of video signal broadcasting.

The measurement system 1 according to the invention is more particularly designed to measure the quality of the images at the output of an image transmission system 10 which may include coding, multiplexing, transcoding, transmultiplexing, modulation and demodulation, respecting the MPEG2 standard.

This system is applicable whenever it is necessary to identify transmission faults in a digital video signal, in particular in order to determine the suitable bit rate for a given sequence of images, as a function of the expected quality.

In FIG. 2, the measurement system 1 according to the invention comprises a synchronization module 11 making it possible to synchronize each coded image to be processed with its coding grid, that is to say the decomposition of the image into blocks of N x N pixels. An example of synchronization processing which can be carried out by this module is described in patent application FR 2 769 452 filed by the Applicant. This synchronization processing in fact makes it possible to locate in the image the decomposition into blocks used by the preceding coding processing carried out by the system 10, and therefore to determine to which coding block By belongs each pixel of the image.

The coded image and its decomposition into blocks are applied to a module 12 for detecting correct macro blocks performing a pre-filtering of correct macro blocks by statistical processing, each macro block of the image grouping together 2 × 2 blocks according to the MPEG2 standard.

The correct macro blocks are discarded from processing, while the others are applied to a module 13 for detecting incorrect macro blocks. The system according to the invention further comprises a module 14 for evaluating an image degradation index which determines a degradation index as a function of the number of erroneous blocks detected by the module 13.

Several types of defects are thus sought either globally, for example the uniform blocks clearly distinct from their vicinity, or locally by considering for example the borders of objects and the textured zones (module 12). We also consider the case where several erroneous blocks are close to each other by comparing the macro blocks to their adjacent blocks, to their diagonal blocks and to the blocks forming the angles of their vicinity (module 13).

In FIG. 3, the module 12 comprises a processing consisting in using the block decomposition defined by the coding grid used during the processing of coding by blocks of the image, and in applying a pre-filtering processing 21 to each macro block of the image, comprising n _M , for example four blocks Bi to B of the coding grid. The purpose of this pre-filtering treatment is to test whether there is equality of two parameters of the macro block considered with those of the n _v = twelve blocks B _v ι to B _v ι ₂ of its neighborhood (see FIG. 4). These parameters are the average luminance and the degree of texture or activity determined by the standard deviation of the luminance.

The average luminance and the standard deviation of each block of the image can be determined directly from the value of the pixels of the image or by applying to the pixels of the image a DCT discrete cosine transform processing. The discrete cosine transform processing consists in applying to each block Bj _j of N x N pixels of the image (N is for example equal to 8) a calculation of the AC and DC transform coefficients obtained using the following formulas : N-1N-1 DC _i = F _iJ (0,0) = - ^ ∑ ∑f _i (x, y) (1)

N _x = 0y = 0 and AC _y (u, v) = F; _j (u, v), with u + v ≠ 0 (2) in which:

fj _j (x, y) represents the luminance of the pixel at point (x, y) in the block Bj _j , x and y being the horizontal and vertical position indices of the pixel respectively the block Bj _j , u and v are between 0 and Nl and represent respectively the indices of the horizontal and vertical spatial frequency, respectively, and c is a function such that c (0) = -η = and c (u) = 1 if u ≠ 0.

The mean luminance μ ^ and the standard deviation σ ^ of luminance are calculated for each block B; J using the coefficients AC and DC of the coding blocks in the DCT domain according to the following formulas:

μi, j = ^DC i, jw and

The two averages compared in module 12 are as follows

'M on the macro block μ _M = ^ μ _B (6) ⁿ M i = ι

on the neighborhood μ _v = - ∑μe> Vι (7) nvi = l

The two standard deviations compared in module 12 are as follows

on the macro block σ _M (8)

on the neighborhood σ _v ( ⁹ )

The equality test is applied in the form of a statistical test, the

Student being used for comparison of means, and Fischer test for comparison of standard deviations (or variances).

According to the Student test, the hypotheses of the equality test are as follows:

H ₀ : μM ⁼ μv against Hi: μ _M ≠ μv (10)

Student's statistic is defined as the ratio of a normal distribution over a law X

or - 2 1 σ _r ((n _M - l) σJι + (n _v - l) σ ^) (12) n _M + n _v - 2 is the combined estimate of variances.

Under the hypothesis H ₀ , the statistic S follows a Student St law (n _M + n _v - 2) with n + n _v - 2 = 14 degrees of freedom. Student's law is tabulated and the threshold corresponding to the risk α chosen.

According to the Fischer test, the hypotheses of the equality test are as follows:

H'o: σ _M ⁼ σ _v against H 'i: σ _M ≠ σy (13)

2

The Fischer statistic is defined as the ratio of two laws of X:

Under the hypothesis H'o, the statistic F follows a Fischer law F (v _1? V ₂ ) where Vi is the degree of freedom of the numerator of the above formula and v ₂ that of the denominator. In other words, Vi is equal to n _M - 1 if G _M > σ _v and to n _v - 1 otherwise.

If the observation of F is less than the value of table F (vι, v ₂ ), we deduce that the standard deviations are equal to the risk α.

It is decided in the processing 22 that the macro block studied corresponds to its neighborhood when these two tests are simultaneously carried out successfully. In fact, this pre-filtering treatment operates by global analysis of the area by extracting in particular the macro blocks which correspond to their vicinity, even in the presence of borders of objects or textured surfaces.

The macro blocks which have not been estimated to be correct by the processing 22 are then applied to three modules 25, 26, 27 applying respective criteria aimed at extracting the macro blocks subject to transmission errors. To this end, each module 25, 26, 27 examines a specific type of fault by applying an appropriate criterion. To this end, the detection of these faults by the modules 25 to 27 is carried out by calculating a fuzzy similarity index obtained in the following manner. We consider the universe Ω of all the blocks of the image to be analyzed and Θ 5 the ordered set of the eight blocks Bi to B ₈ neighboring a block B ₀ considered. The similarity of block B ₀ with its neighborhood Θ is defined as a fuzzy set A / Θ in the universe Ω. The membership function Λ _{A / Θ} associated with the fuzzy set A / Θ is a function of the universe Ω towards the set [0, 1], 0 corresponding to a low similarity and 1 corresponding to a high similarity. In the case where the blocks are considered to be characterized by their average luminance (or the average of their gray levels), two blocks are similar if the difference between their average luminances is small. This pairwise similarity between two blocks X and Y of the set considered is approximated by the membership function μ _{c! Ose} defined on the set Ω x Ω by the following formula: is μ _c iose (> Y) = ι - ^ | μχ - μγ | (15) in which M = 255 and corresponds to the maximum value admissible by the eight differences of the mean of the central block B ₀ and the means of the neighboring blocks Bi to B ₈ , and μ _x and μ _γ are the means of the levels of gray (or average luminance) of blocks X and Y. 0 The global similarity index between block B ₀ and the blocks in its vicinity is given by:

Λ _{A Θ} (B ₀ ) = Max (μ _c ι _ose (B ₀ , Bj), (B ₀ , B ₂ ), ..., (B ₀ , B ₈ ) (16)

The overall similarity index obtained is then compared to a threshold s which 5 depends on the texture level of its vicinity and which is obtained by the following formula: s = exp (-aσ) or e ^"aσ (17) in which σ is the standard deviation evaluated on block B ₀ and its neighborhood, and a is a positive constant determined experimentally. This formula allows 0 to perform an adaptive filtering aimed at taking into account a masking effect. The criterion is considered verified ( erroneous block) if the index α is less than the threshold s.

The module 25 performs an adjacent similarity measurement by comparing each pair of adjacent blocks of each macro block with the adjacent pair of blocks in the vicinity of the macro block. Thus, the module performs for each macro block not excluded by the pre-filtering processing 21, the average luminance comparisons of the following pairs of blocks (see numbering of the blocks in FIG. 4):

- (B ₁ , B ₂ ) and (B _v2 , B _v3 ),

- (B ₂ , B ₄ ) and (B _v6 , B _v8 ),

- (B ₃ , B ₄ ) and (B _vl0 , B _vll ), and

- (B ₃ - Bi) and (B _v5 , B _v7 ).

The comparison algorithm calculates the differences in the average luminances of these pairs of block sets:

Δ ¹ , (μβ, + μB.) - (μβ _v V2. + Μβ> _vV3 ,) (18)

Δ ' ₂ = (μ _B , + μB.) - (μ _Bv + μ _B > _v V ₈ (19)

Δ3 = - | (μ _B3 + μB ₄ ) - (μB _vl0 + μB _vπ ) (20)

^Δ 4 = 2 | (μβ ₃ + μβ,) - (μβ _V5 + μβ _V7 ) (21)

Then, it only keeps the second maximum Δ 2max of the four calculated difference values, to allow a significant difference in one direction:

4 _ma x = 2 ^nd max (Δ ^I ₁ , 4, ^ ₃ , ^ ₄ ,) (22)

Then, in order to obtain a fuzzy similarity index α, it applies to this second maximum value the function μ _c ι _ose : ^α cIosev ^ 2max ^ _* r 2max (23)

This index α is then compared to a threshold Si obtained using the formula (17) in which in which the standard deviation σ is calculated over the entire study region comprising the macro block (Bi, B ₂ , B ₃ , B ₄ ) and all neighboring blocks B _v j to B _vl2 , and the parameter a is equal to 1/12. The macro block studied is considered erroneous by the module 25 if the index α is less than the threshold if.

In summary, the module 25 does not consider as erroneous the macro blocks situated along a frank border of an object of the image. Macro blocks selected as erroneous must have at least two borders with adjacent blocks along which its mean luminance is sufficiently distinct from that of the adjacent blocks.

The module 26 performs for each macro block a measurement of diagonal similarity by comparing the average luminance of each corner block of the macro block (Bi, B ₂ , B ₃ , B ₄ ) with the blocks B _v ι, B _v4 , B _v9 and B _vl2 respective neighbors, located on their diagonal. The purpose of this measurement is to detect erroneous adjacent macro block bands by a perception of non-similarity with diagonal macro blocks.

To this end, the module 26 evaluates for each macro block not removed by the pre-filtering processing 21, four differences in average luminance between each of the blocks of the macro block and the respective diagonal blocks in the vicinity of the macro block:

He then selects the second minimum Δ 2 min of these values because we are looking for dissimilarity.

Δ? _min = 2 ^nd min (Δy, Δ5, Δ?, Δ5,) (28)

Then, in order to obtain a fuzzy similarity index β, it applies to this second minimum value the function μ _c ι _ose defined by the formula (23): β = μclose (Δ2min) (29)

The value obtained β is then compared to a threshold s ₂ calculated by formula (17) in which the standard deviation is calculated over the entire study region including the macro block (B], B ₂ , B ₃ , B ₄ ) and all neighboring blocks B _v ι to B _v ι ₂ and the parameter a is equal to 1/12. If the value obtained is less than the threshold s ₂ , the macro block is considered to be erroneous.

The criterion applied by module 26 thus makes it possible to determine that the macro block considered is erroneous if at least three diagonal blocks do not have a similar average luminance, which allows an object border to be authorized in any diagonal.

The module 27 performs an angle similarity measurement by comparing the average luminance of each angle block (Bi, B ₂ , B, B ₄ ) of each macro block not removed by the pre-filtering treatment 21 with the average luminance of each of the three adjacent blocks in the vicinity of the macro block, delimiting the angles of the macro block, namely B] with (B _v5 , B _v B _v2 ), B with (B _v3 , B _v4 , B _v6 ), B ₄ with ( B _v8 , B _vl2 , B _vn ) and B ₃ with (B _v ι ₀ , B _v9 , B _v7 ). This comparison aims to combine the effects observed at the intersection of several macro blocks in the neighborhood.

To this end, for each set of neighboring angle blocks, the module 27 calculates the differences in average luminance between each block of the set of angle blocks and the adjacent block of the macro block considered, and selects the minimum value of the three values obtained:

= min (| μ _B μβ VI μβ, - μβ V2 μβ, - μβ, (30)

4 "min (| μ _B2 - μ _By4 μβ. - μβ V ₃ μβ ₇ - μβ (31) ^Δl ₃ ^{II =} min (| μ _B3 - μ _By9 μ ^β ₃ μβ V7 μβ. Μβ (32) Δ ₄ ^π = min (| μ _B4 - μ _Byi2 μ ^β . μβ V ₈ μ ^β _ - μβ VI, ι ⁾ (33)

We thus obtain four difference values (one per angle of the macro block), among which we select the second maximum value Δ 2max to tolerate a frank object border in any direction: Aψ _max = 2 ^nd max (Δ ^I ₁ ^π , Δ ?, Δ ₃ ^π , Δ ¹ ] ¹ ,) (34)

The function μ _c ι _ose defined by formula (23) is then applied to the value retained:

Y = μclose (Δ2 ^π max) (35)

The fuzzy similarity index γ obtained is compared with a threshold s ₃ which is calculated using formula (17) in which the standard deviation is calculated over the entire study region including the macro block (Bi, B ₂ , B ₃ , B ₄ ) and all the neighboring blocks B _v ι to B _v i ₂ and the parameter a is chosen equal to 1/20. The value of the parameter a is adapted so that the number of macro blocks selected by the module 27 is of the same order as those selected by modules 25 and 26.

A macro block is then considered as erroneous by the module 27 if the index γ is less than the threshold s ₃ , that is to say when at least two of its corner blocks are not similar to the blocks of adjacent corner of its vicinity.

The macro blocks detected in error by the modules 25, 26 and 27 are then applied to a module 28 for deciding on the state of each macro block, as a function of the state of the macro block determined by each of the modules 25 to 27. If the macro block is detected as erroneous by at least two of these modules, it is considered erroneous by the module 28.

The erroneous blocks are then transmitted to the module 14 for evaluating an image degradation index which accounts for the number of macro blocks thus considered erroneous in the image, and determines from this counting a degradation index of l image obtained by dividing the number of erroneous macro blocks by the total number of macro blocks in the image.

Comparative tests of the method for measuring transmission errors which has just been described with those of the prior art, make it possible to show that this method is much more effective in reducing the risk of false detection, while remaining sufficiently reactive to detect erroneous image areas.

This method is particularly effective in detecting critical points in an image transmission chain, which then allows the service provider to intervene to improve the quality of the service as soon as damage occurs to the end user. It thus contributes to optimizing the use of transmission resources.

In the process which has just been described, the criteria for selecting the erroneous macro blocks are applied to the average luminance of the blocks of the image. It is of course possible to envisage applying these criteria to other parameters of the blocks of the image, determined on the basis of the value of the pixels of the image.

Claims

1. Method for measuring degradations of a digital video image introduced by a transmission system (10), the image being coded using a coding process involving a decomposition of the image into blocks and a calculation of transform by blocks, the image being transmitted by macro blocks grouping together several blocks, characterized in that it comprises steps consisting in:

- perform a pre-filtering (12) consisting in applying to each macro block of the transmitted image a statistical test to determine if there is equality between a property of the macro block and a corresponding property of the blocks adjacent to the macro block, and retain the macro blocks whose property is not equal to that of its adjacent blocks,

- for each macro block selected, determine (13) whether the blocks of the macro block have a property similar, according to several direction criteria, to that of neighboring blocks, and consider that the macro block is erroneous if it is determined not similar according to several of the management criteria, and

- determining (14) an image degradation index as a function of the number of macro blocks considered to be erroneous in the image.

2. Method according to claim 1, characterized in that the direction criteria comprise a first criterion (25) consisting in comparing a property of each macro block selected with that of neighboring blocks having a common border with the macro block, a second criterion (26) consisting in comparing a property of each macro block retained with that of neighboring blocks situated on diagonals of the macro block, and a third criterion (27) consisting in comparing a property of each macro block retained with that of neighboring blocks delimiting the angles of the macro block, a macro block being considered erroneous if it is determined not similar according to at least two of the three criteria.

3. Method according to claim 1 or 2, characterized in that it comprises a prior synchronization step (11) consisting in determining the coding grid of the coded image, in order to find the decomposition into coding blocks of the image, used when coding the image.

4. Method according to one of claims 1 to 3, characterized in that the determination (13) of the similarity of the property of a macro block of the image with that of the neighboring blocks consists in determining a fuzzy similarity index and in comparing the determined similarity index with a dependent threshold the value of the pixels of the image in an area in which the macro block is located.

5. Method according to claim 4, characterized in that the fuzzy similarity index is obtained using a membership function applied to an average luminance difference between the macro block and neighboring blocks of the macro block.

6. Method according to claim 5, characterized in that the membership function is defined by the following formula: μ _close (Δ) = l- ^ | Δ |

7. Method according to one of claims 1 to 6, characterized in that a macro block is considered to be erroneous according to a first criterion (25) if it has at least two borders along which the blocks of the macro block have a Average luminance sufficiently different from that of adjacent blocks in the vicinity of the macro block.

8. Method according to one of claims 1 to 7, characterized in that a macro block is considered to be erroneous according to a second criterion (26) if at least three of the corner blocks of the macro block have a sufficiently different mean luminance of that of respective diagonal blocks in the vicinity of the macro block.

9. Method according to one of claims 1 to 8, characterized in that a macro block is considered to be erroneous according to a third criterion (27) if at least two corner blocks of the macro block have an average luminance sufficiently different from those of blocks in the vicinity of the macro block, adjacent to the respective corner block.

10. Method according to one of claims 1 to 9, characterized in that the pre-filtering (12) consists in applying to each macro block of the image a statistical equality test to determine if there is equality of average luminance between the macro block and the neighboring blocks of the macro block.

11. System for measuring degradations of a digital video image introduced by a transmission system, characterized in that it comprises means for implementing the method according to one of the preceding claims.