T 1616/06 () of 16.11.2010

Summary of Facts and Submissions

I. This appeal is against the decision of the examining division to refuse European patent application No. 98 906 493.6, published as WO 98/36553 A2.

II. The decision under appeal was based on the grounds that the subject-matter of claims 1, 7 and 9 of the main and second auxiliary requests did not involve an inventive step (Article 56 EPC 1973) and that the first auxiliary request was not admitted into the procedure in accordance with Rule 86(3) EPC 1973.

III. With the statement of grounds of appeal, the co-appellants (hereinafter "the appellants") submitted complete documents of a main request, first and second auxiliary requests, respectively, including pages 1, 2, 2A and 3 to 14 of the description and drawings sheets 1/4 to 4/4.

IV. In a communication accompanying the summons to oral proceedings the board, inter alia, expressed doubts as to whether claim 1 according to the main request was clear and supported by the description (Article 84 EPC 1973).

V. With a letter dated 13 October 2010 the appellants filed respective sets of amended claims according to a main request and an auxiliary request, replacing all requests then on file, as well as the following document as supporting evidence for the technical meaning of certain terms used in claim 1:

D9: P. Symes, "Digital Video Compression", McGraw-Hill, 2004, cover sheets and pages 67-87

VI. Oral proceedings were held before the board on 16 November 2010. During the oral proceedings, the appellants filed a set of amended claims according to a second auxiliary request.

VII. The appellants' final requests are that the decision under appeal be set aside and that a patent be granted on the basis of the main request, or, in the alternative, the first auxiliary request, both filed with a letter of 13 October 2010, or the second auxiliary request, filed in the oral proceedings.

VIII. Independent claim 1 according to the main request reads as follows:

"A method for adjusting decoded pixel values that are representative of an image, where said decoded pixel values have previously undergone a transform operation and a quantization operation, said method comprising the steps of:

(a) selecting a block of decoded pixel values;

(b) dividing said block into a plurality of subblocks;

(c) computing an adjustment value for each of said subblocks;

(d) applying a forward transform to the plurality of adjustment values corresponding to said plurality of subblocks to produce a plurality of transform coefficients;

(e) determining whether said plurality of adjustment values should be applied to said decoded pixel values, said determining step comprising comparing said plurality of transform coefficients with a quantization step employed by said quantization operation; and

(f) adjusting said decoded pixel values within each subblock in accordance with said plurality of adjustment values, if said determining step (e) determines that said plurality of adjustment values should be applied, and if said determining step (e) determines that said plurality of adjustment values should not be applied to a given subblock, outputting said given subblock and repeating steps (c) through (f) for remaining subblocks in said plurality of subblocks."

Claims 2 to 9 according to the main request have no bearing on the present decision.

IX. Independent claim 1 according to the first auxiliary request has the same wording as claim 1 of the main request except for feature (c):

"…

(c) computing an adjustment value for each of said subblocks, said adjustment value being configured to adjust the median pixel value of each of said plurality of subblocks;

…"

Claims 2 to 15 according to the first auxiliary request have no bearing on the present decision.

X. Independent claim 1 according to the second auxiliary request reads as follows:

"A method for adjusting decoded pixel values that are representative of an inter-coded image, where said decoded pixel values have previously undergone a transform operation and a quantization operation, said method comprising the steps of:

(a) selecting a block of decoded pixel values corresponding to the block size of the transform used in the encoder;

(b) dividing said block into a plurality of subblocks;

(c) computing an adjustment value for each of said subblocks, said adjustment value being configured to adjust the median pixel value of each of said plurality of subblocks to have a value of zero;

(d) applying a forward transform to the plurality of adjustment values corresponding to said plurality of subblocks to produce a plurality of transform coefficients, the forward transform being the same as the transform operation previously applied to the pixel values;

(e) determining whether said plurality of adjustment values should be applied to said decoded pixel values, said determining step comprising comparing said plurality of transform coefficients with a quantization step employed by said quantization operation;

(f) adjusting said decoded pixel values within each subblock by applying the respective adjustment value to each pixel value of the subblock, if said determining step (e) determines that said plurality of adjustment values fit within the quantization step, and if said determining step (e) determines that said plurality of adjustment values for a given subblock do not fit within the quantization step, outputting said given subblock without adjustment and repeating steps (c) through (f) for remaining subblocks in said plurality of subblocks; and

(g) if after repeating steps (c) through (f) until no sub-blocks remain, no subblock adjustment values are computed for which the corresponding transform coefficient fits within the quantization step, no adjustment values are applied to any subblock of the selected block."

[The amendments over claim 1 according to the first auxiliary request (essentially added features) are set in italics.]

Claims 2 to 9 according to the second auxiliary request have no bearing on the present decision.

XI. The appellants argued essentially as follows regarding the requirements of Article 84 EPC 1973 for claim 1 according to the main and first auxiliary requests and regarding the admissibility of the second auxiliary request:

Main request - Article 84 EPC 1973

Re "adjustment value"

While the board has expressed doubts as to the clarity of the term "adjustment value" in the context of claim 1, the board seems to express a clear and unambiguous interpretation of this term. The appellants agree with the interpretation of this term as presented by the board as this appears to be entirely appropriate in the context of the presently claimed invention. Specifically, the appellants believe that the board's interpretation of "adjustment value" as "a value with certain limitations which are not clearly defined" to be, for all practical purposes, the same as "a value". Thus, the term is clear in the sense required by Article 84 EPC 1973.

Re "forward transform" and "plurality of transform coefficients"

The appellants disagree that the terms "forward transform" and "plurality of transform coefficients" in step (d) of claim 1 lack clarity for the following reasons.

D9, a textbook on Digital Video Compression, describes the term "transform" at page 68, first paragraph; the reversibility of transforms at page 69, last paragraph; the term "forward transform" at page 82, paragraphs 1 and 4; the production of coefficients using transforms from the last paragraph on page 70 through the first full paragraph on page 71, page 75 in its entirety, page 78, first full paragraph, page 80, lines 4 and 5, page 80, first full paragraph, and page 87, second full paragraph; and the term "transform coefficients" at page 78, first full paragraph, and page 87, second full paragraph.

Furthermore, the claim clearly defines the role played by the plurality of transform coefficients as being part of the determining step (e) and the role of the forward transform as being to produce the transform coefficients. These uses of the terms and roles of the claim elements are entirely consistent with the knowledge, understanding and/or use of the terms by those skilled in the art. Thus it is submitted that the present claim language "forward transform" and "plurality of transform coefficients" is clear and intelligible to the skilled person, in accordance with Article 84 EPC 1973.

Re Step (e)

Although the appellants do not fully agree with the board's concern, that determining step (e) is merely defined as a result to be achieved and therefore not clear, steps (e) and (f) in claim 1 have been amended to address the board's concern. Specifically, step (e) of claim 1 has been amended to specify that the determining step comprises comparing a plurality of transform coefficients with a quantization step employed by the quantization operation that the decoded pixel values have previously undergone. Also, step (f) of claim 1 has been amended to recite the outcome where the determining in step (e) reaches a negative conclusion. Thus it is submitted that step (e) is clear in the context of the claim as a whole, in accordance with Article 84 EPC 1973.

First auxiliary request - Article 84 EPC 1973

In view of the board's opinion on the clarity of the term "adjustment value", the claims have been amended to indicate a technical definition of this feature relating to other ones of the claim elements. Thus further technical meaning is imparted to this term such that the claims are clear in accordance with Article 84 EPC 1973.

Admissibility of the second auxiliary request

The second auxiliary request should be admitted into the proceedings because the amendments made to claim 1 according to this request overcome the board's objections under Article 84 EPC 1973. These amendments could not have been filed earlier because the appellants did not have all the details of the board's objections before the oral proceedings.

Reasons for the Decision

1. The appeal is admissible.

2. Although D9, to which the appellants referred in support of the technical meaning of certain terms, was published more than five years after the priority date of the present application, the board considers that the relevant passages of this extract of a textbook on digital video compression do not go beyond the common general knowledge at the priority date. The board will thus also refer to this textbook extract below.

Main request - Article 84 EPC 1973

3. The method of claim 1 concerns a "method for adjusting decoded pixel values that are representative of an image".

According to the description of the application, the technical problem solved by the invention is to reduce blocking and ringing artifacts associated with the coding of images or video frames (see page 1, lines 8 to 10). The description further states that the method of the invention solves this problem by recovering or adjusting quantized coefficients by using a nonlinear scheme (see page 1, lines 10 to 12, and page 2A, lines 4 and 5, corresponding to page 2, lines 5 to 7, of the published application).

4. Re the term "adjustment value[s]"

Claim 1 does not define the term "adjustment value[s]" in steps (c) to (f). The board does not dispute the appellants' assertion, supported by a printout of an internet consultation of the Compact Oxford English Dictionary filed before the examining division, that the verb "to adjust" means "to alter slightly so as to achieve a desired result". The board, however, does not regard the expression "adjustment value" as having a clear meaning in the context of claim 1 for the following reasons. Firstly, in the absence of more information in the claim, it is ambiguous whether the "adjustment value" represents the difference between an initial value and an adjusted value, or the adjusted value itself. Secondly, claim 1 does not indicate the desired result of the slight alterations in the meaning of the dictionary definition. Nor does claim 1 clearly indicate how the adjustment values are calculated and to what effect. Whereas claim 1 specifies that the plurality of forward transformed adjustment values (one for each of a plurality of subblocks) is used for a comparison with "a quantization step" for determining whether the plurality of adjustment values should be applied or not, neither the grouping of values nor the comparison (individually or groupwise) help to clarify the purpose of the adjustments. In the absence of such information the expression "computing an adjustment value" could mean anything ranging from "computing a value" to "computing a value with certain limitations not clearly defined".

The appellants argued that the term "an adjustment value" means, for all practical purposes, "a value".

The board cannot share this view because it is based on the assumption that the word "adjustment" can simply be ignored as having no limiting effect whatsoever. This assumption, however, is not correct because, as the above dictionary definition of the verb "to adjust" indicates, the word "adjustment" must have a limiting effect, both as limiting the range in which the value may be ("to alter slightly") and as to the direction of the alteration ("to achieve a desired result"). Both the range and the direction appear important in steps (d) to (f) of claim 1 in order to determine whether the plurality of adjustment values should be applied or not. Moreover, the board is not convinced that the description of the application supports an interpretation of the term "adjustment value" as broad as put on by the appellants.

5. Re the term "forward transform"

Claim 1 does not provide any information as to the type of the "forward transform". A "transform", according to the definition given on page 68 of D9 is "a mathematical rule that may be used to convert one set of values to a different set". The word "forward" before "transform" is commonly used to distinguish the transform from its mathematical reciprocal called the "inverse transform". The expression "forward transform" in claim 1 thus covers any type of mathematical transform. By way of contrast, the description states that the forward transform is "equivalent to the forward transform that was used in the encoder to encode the original signal" (see page 2A, lines 8 to 11, of the description). Such forward transforms used for encoding blocks of an image are known to be transforms from the time or spatial domain to the frequency or spatial frequency domain, such as a Discrete Cosine Transform (DCT) or a wavelet transform (see page 4, lines 13 to 16, of the description). The expression "forward transform" in claim 1 is therefore much more general than the specific types of transforms referred to in the description. The application does not disclose how the technical problem of reducing blocking and ringing artifacts associated with the coding of images or video frames could be solved without the "forward transform" of claim 1 being a transform from the spatial domain to the spatial frequency domain, in particular the transform which was used for encoding where the problem was created in a lossy quantization step (see page 1, lines 15 to 34, of the description). The board thus considers that there is a lack of support in the description for the use of such a broad expression as "forward transform" without further limitations.

The appellants referred to passages of D9 to support their argument that the expression "forward transform" in claim 1 has a clear meaning.

Even if, contrary to what can be understood from the description, it were not necessary in the context of claim 1 that the forward transform is equivalent to the one used in the encoder, it is still necessary to distinguish between the different meanings this expression may have. Pages 68 and 69 of D9, describe "transforms" in general, and pages 70 to 87 of D9 refer to a specific class of transforms from the time or spatial domain to the frequency or spatial frequency domain, in particular the Fourier transform, and more particularly the Discrete Cosine Transform (DCT). The first paragraph on page 68 of D9 confirms the board's view that a transform also covers "a mathematical rule that may be used to convert one set of values to a different set". The passages cited by the appellants on pages 70 to 87 apply to Fourier transforms, in particular to DCT transforms. The expression "forward transform" in claim 1, however, is not limited to these types of transforms. The appellants' arguments are based on an unduly narrow meaning of this expression which also finds no support in D9. Hence the appellants' arguments fail to convince the board.

Therefore, the board concludes that the expression "forward transform" has a broad meaning which leads to a lack of support in the description in the present case, as explained above.

6. Moreover, step (d) compounds the problems discussed in sections 4 and 5 supra by stating that the "forward transform" is applied to the "adjustment values". This is so because the comparison with a quantization step is made with the transform coefficients which evidently depend on the type of the forward transform.

7. Conclusion

For the above reasons, claim 1 does not fulfil the requirements of Article 84 EPC 1973. Thus the appellants' main request is not allowable.

First auxiliary request - Article 84 EPC 1973

8. Claim 1 according to the first auxiliary request adds the following phrase at the end of step (c): "said adjustment value being configured to adjust the median pixel value of each of said plurality of subblocks".

This additional phrase essentially states that the adjustment value "adjusts" the median pixel value of each subblock. However, the purpose of the adjustment remains unclear as in claim 1 according to the main request. Whereas claim 1 now specifies that median adjustment values are computed for each subblock (one value per subblock), the effect of the comparison of the "plurality of transform coefficients" (i.e. the transformed median adjustment values) with a quantization step remains unclear. The same is true for the type of the forward transform. The description appears to make an assumption of median values around zero and to make adjustments so that the medians will be caused to be zero (see for instance page 10, lines 5 to 20, and figure 2: 220 to 250). However, the board does not see support in the description for a comparison of the plurality of (median) transform coefficients with a quantization step in general, in particular where the forward transform is not necessarily equivalent to the one used in the encoder. Hence the above additional wording does not overcome the objections under Article 84 EPC 1973 raised under sections 4 to 6 supra.

Accordingly, the appellants' first auxiliary request is not allowable.

Second auxiliary request - Article 13 RPBA

9. According to Article 13(1) RPBA (Rules of Procedure of the Boards of Appeal, OJ EPO 2007, 536), any amendment to a party's case after it has filed its grounds of appeal may be admitted into the proceedings and considered at the board's discretion. The discretion shall be exercised in view of inter alia the complexity of the new subject-matter submitted, the current state of the proceedings and the need for procedural economy.

10. In the present appeal proceedings, the appellants filed amended claims 1 to 9 during the oral proceedings before the board. These claims comprised numerous extensive amendments compared to the claims then on file or previously considered during the appeal proceedings. Some of the amendments were indeed an attempt to overcome objections which the board had raised in the communication annexed to the summons to oral proceedings and further explained in oral proceedings. However, at a very late stage of the appeal proceedings, these amendments also comprised features which had never been discussed before. These new features obviously raised new and complex issues under Article 84 EPC 1973 which needed further examination and discussion. For instance, one of the amendments to claim 1 referred to a "plurality of adjustment values for a given subblock", which contradicts step (c) of the claim stating that there is one adjustment value for each subblock. The presentation of complex subject-matter at such a late stage of proceedings goes, however, against procedural economy.

For the above reasons, the board exercised its discretion under Article 13(1) RPBA and did not admit the appellants' second auxiliary request into the proceedings.

Conclusions

11. Since the appellants' main and first auxiliary requests are not allowable and the appellants' second auxiliary request was not admitted into the proceedings, the appeal must be dismissed.

ORDER

For these reasons it is decided that:

The appeal is dismissed.