Key points
- The Board in the headnote: “there is no need that a prior art document explicitly mentions the claimed features. It is necessary and sufficient that an embodiment falling under the claim scope be directly and unambiguously derivable from the prior art document. ”
- This summary of the rule regarding novelty seems correct to me. See also my article in epi Information 4/2020.
- The Board, in the reasons of the decision:
“The Board agrees that the same yardstick should be used [for added matter, i.e. Art.123(2) and novelty]. However, this yardstick does not require an explicit mention of the relevant features, but only that the features be directly and unambiguously derivable. It is also not necessary that the exact claim wording, considering its breadth and all its possible interpretations, be derivable from a prior art document for the claim to lack novelty. It is necessary and sufficient that an embodiment falling under the claim scope be derivable from the prior art document.” - The Board then applies this rule to the case at hand: “The Board considers this to be the case here, because the skilled person is told to use conventional fingerprint images for identification, which directly leads to the necessity of storing these images in an organized way, i.e. to a database of images, although this is not explicitly mentioned.” The Board then hence to admit that an alternative implementation of the prior art document E2 is possible: “One may of course also (or instead) store feature values extracted from the images in a database.”
- The Board then reasons that: “the fact that an alternative embodiment exists cannot change the fact that the first one is directly and unambiguously derivable. In fact, it is possible that multiple alternatives can be considered directly und unambiguously derivable, even when none is explicitly mentioned.” The Board summarizes this in the headnote: “That an alternative exists does not change this: it is possible that multiple alternatives can be considered directly und [sic] unambiguously derivable, even when none is explicitly mentioned.”
- I'm not sure if this is established case law. If an embodiment is not explicitly mentioned, it must be implicit disclosure for lack of novelty. In principle, multiple alternative embodiments can each be implicitly disclosed. There is however some case law setting a very high bar for implicit disclosure.
- I also note that the Board be reasoning as it did, could leave open whether cited patent application E2 was prior art under Art.54(2) or a prior right under Art.54(3), which was a matter of dispute.
T 1599/18 -
https://www.epo.org/law-practice/case-law-appeals/recent/t181599eu1.html
Auxiliary request 2
11. In view of the analysis above as to the novelty of the main request, the Board agrees with the decision of the Opposition Division that claim 1, in the alternative under a), is anticipated by a combination of claims 30 and 32 of E2. The appellant did not contest this assessment.
12. The Board also agrees with the Opposition Division that alternative b) is disclosed by paragraph 72 in E2 which teaches to extract a conventional fingerprint image for biometric assessment.
12.1 Paragraph 72 of E2 teaches that "a conventional fingerprint image is extracted from the sensor for further processing and biometric assessment".
12.2 In E2 it is made clear to the skilled person that one biometric assessment to be performed is identification; in fact this can be considered to be the main purpose of the invention (see paragraphs 4, 13, 42); this is otherwise the normal use of a "conventional fingerprint image". Thus E2 directly and unambiguously teaches the extraction of conventional fingerprint images for identification.
12.3 Identification with conventional fingerprint images implies a comparison of the extracted conventional fingerprint image with some conventional fingerprint images of individuals. These conventional fingerprint images of individuals constitute, in the words of the claim, "a plurality of non-multispectral images stored in a database relating the plurality of non-multispectral images to identities of people". It is noted that the word "database", in the broadest sense, means nothing more than an organized collection of data.
13. The appellant contested that one can unambiguously derive from E2 the usage of a database with a plurality of non-multispectral images. There was no reference to such a database. One should use the same yardstick for added matter and novelty. The addition of this wording would never be allowed in E2.
14. The Board agrees that the same yardstick should be used. However, this yardstick does not require an explicit mention of the relevant features, but only that the features be directly and unambiguously derivable.
14.1 It is also not necessary that the exact claim wording, considering its breadth and all its possible interpretations, be derivable from a prior art document for the claim to lack novelty. It is necessary and sufficient that an embodiment falling under the claim scope be derivable from the prior art document. The Board considers this to be the case here, because the skilled person is told to use conventional fingerprint images for identification, which directly leads to the necessity of storing these images in an organized way, i.e. to a database of images, although this is not explicitly mentioned.
14.2 One may of course also (or instead) store feature values extracted from the images in a database. But the fact that an alternative embodiment exists cannot change the fact that the first one is directly and unambiguously derivable. In fact, it is possible that multiple alternatives can be considered directly und unambiguously derivable, even when none is explicitly mentioned.
Regarding the headnote “That an alternative exists does not change this: it is possible that multiple alternatives can be considered directly und [sic] unambiguously derivable, even when none is explicitly mentioned.”
ReplyDeleteMaybe it can be seen as a selection from a single list, which does not confer novelty?