Ex Parte Ratner et alDownload PDFPatent Trial and Appeal BoardMay 21, 201311821767 (P.T.A.B. May. 21, 2013) Copy Citation UNITED STATES PATENT AND TRADEMARKOFFICE UNITED STATES DEPARTMENT OF COMMERCE United States Patent and Trademark Office Address: COMMISSIONER FOR PATENTS P.O. Box 1450 Alexandria, Virginia 22313-1450 www.uspto.gov APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO. 11/821,767 06/25/2007 Edward R. Ratner 10043.000410 7452 31894 7590 05/21/2013 OKAMOTO & BENEDICTO, LLP P.O. BOX 641330 SAN JOSE, CA 95164 EXAMINER MOTSINGER, SEAN T ART UNIT PAPER NUMBER 2669 MAIL DATE DELIVERY MODE 05/21/2013 PAPER Please find below and/or attached an Office communication concerning this application or proceeding. The time period for reply, if any, is set in the attached communication. PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE ____________________ BEFORE THE PATENT TRIAL AND APPEAL BOARD ____________________ Ex parte EDWARD R. RATNER and SCHUYLER A. CULLEN ____________________ Appeal 2010-007698 Application 11/821,767 Technology Center 2600 ____________________ Before: DEBRA K. STEPHENS, LYNNE E. PETTIGREW, and STACEY G. WHITE, Administrative Patent Judges. STEPHENS, Administrative Patent Judge. DECISION ON APPEAL Appeal 2010-007698 Application 11/821,767 2 STATEMENT OF CASE Appellants appeal under 35 U.S.C. § 134 from a final rejection of claims 2-4, 6-8, 11-12, and 15-16. We have jurisdiction under 35 U.S.C. § 6(b). Claims 1, 5, 9-10, and 13-14 have been canceled.1 We AFFIRM. Introduction The claims are directed to a computer-implemented method for automated object recognition and classification in scenes using segment- based object extraction (Abstract). STATEMENT OF THE CASE Exemplary Claim Claim 2, reproduced below, is illustrative of the claimed subject matter (disputed limitations in italics): A computer-implemented method for automated image object recognition and classification, the method comprising: receiving a video comprising a sequence of static image frames; 1 Appellants canceled claims 1, 5, 9, 10, 13, and 14 in their Appeal Brief. (App. Br. 2). Although the Examiner’s Headings in Grounds of Rejection (Ans. 3, §Grounds of Rejection) do not omit the canceled claims, the Examiner does indicate the amendment filed with the brief has been entered and made of record (Ans. 2, §Status of Amendments After Final). Further, the Examiner states the statement of the status of claims contained in the brief is correct (Ans. 2, §Status of Claims) and omits the canceled claims from the body of the Grounds of Rejection. Appeal 2010-007698 Application 11/821,767 3 creating a sequence of segmentation maps corresponding to the sequence of static image frames, wherein each segmentation map divides an image frame into segments; grouping segments so as to form extracted objects which move from frame to adjacent frame in the video; extracting features from said extracted objects; and classifying said extracted objects using the features, wherein extracting features from an extracted object comprises: loading a pixel mask of the extracted object; and selecting keypoints using a keypoint selection technique which is applied only to pixels belonging to the extracted object. REFERENCES Karmann Courtney US 5,034,986 US 5,969,755 Jul. 23, 1991 Oct. 19, 1999 Zhang US 2002/0169532 A1 Nov. 14, 2002 REJECTIONS The Examiner made the following rejections: (1) Claims 2-4 and 6-8 stand rejected under 35 U.S.C. §103(a) as being unpatentable over Karmann and Zhang. Appeal 2010-007698 Application 11/821,767 4 (2) Claims 11, 12, 15, and 16 stand rejected under 35 U.S.C. §103(a) as being unpatentable over Karmann, Zhang, and Courtney. We have only considered those arguments that Appellants actually raised in the Briefs. Arguments Appellants could have made but chose not to make in the Briefs have not been considered and are deemed to be waived. See 37 C.F.R. § 41.37(c)(1)(vii)(2011). ISSUE 1 35 U.S.C. § 103(a): claims 2-4 and 6-8 Appellants assert their invention is not obvious over Karmann and Zhang because neither reference teaches “selecting keypoints using a keypoint selection technique which is applied only to pixels belonging to the extracted object” (App. Br. 5-7). Specifically, Appellants contend the cited portion of Zhang merely discloses a five parameter definition of an ellipse that encloses a portion of a silhouette (App. Br. 7). According to Appellants, a majority of the points on the ellipses lie outside of the silhouettes and thus, the points on the ellipses as well as the centerpoint of the ellipse do not belong to the silhouette (id.). Further, Appellants argue one of ordinary skill in the pertinent art would know: A silhouette is merely a binary mask of an object, not an extracted object and The claimed keypoints are used to define features of an object to classify an object, and an ellipse center is insufficient for such a classification. Appeal 2010-007698 Application 11/821,767 5 (Reply Br. 2). Issue 1: Has the Examiner erred in finding the combination of Karmann and Zhang teaches or suggests “selecting keypoints using a keypoint selection technique which is applied only to pixels belonging to the extracted object” as recited in independent claims 2 and 6? ANALYSIS Appellants’ arguments are not persuasive. Specifically, we have reviewed the Examiner’s rejection in light of Appellants’ arguments that the Examiner has erred. We adopt as our own the findings and reasons set forth by the Examiner in the action from which this appeal is taken and the reasons set forth by the Examiner in the Answer in response to Appellants’ Appeal Brief. However, as outlined infra, we highlight and address specific findings and arguments regarding claim 2 for emphasis. Appellants have not explicitly defined the terms “extracted object” or “keypoints” in their Specification. Nor have Appellants presented sufficient evidence or argument to show error in the Examiner’s interpretations. Appellants argue “one of ordinary skill in the pertinent art would know a silhouette is merely a binary mask of an object, not an extracted object” (Reply Br. 2); however, Appellants present no evidence to support that assertion. Similarly, Appellants argue “one of ordinary skill in the pertinent art would know that, while the claimed keypoints are used to define features of an object to classify an object (determine what type of object it is), an ellipse center per Zhang is insufficient for such classification” (Reply Br. 2). Again, Appellants present no evidence to support that assertion. Appeal 2010-007698 Application 11/821,767 6 With respect to Zhang, we agree with the Examiner that the keypoints are the center points of the ellipse (Ans. 11). We also agree with the Examiner that the Appellants’ claim does not require the keypoints be part of the object - it only requires the method used to select keypoints is applied only to pixels belonging to the extracted object (Ans. 13). Thus, we agree selected keypoints can be inside or outside of the object. We further agree Zhang discloses an ellipse fitting technique which is applied only to the silhouette extracted from the object (Zhang [0023]). We also agree with the Examiner that since the ellipse is determined by fitting an ellipse to a segment of the object, only pixels that are part of the object are used when fitting the ellipse (Ans. 14). The pixels that are not part of the segmented object are “set to zero” and ignored (Ans. 13; Zhang, [0022] and [0023]). We also agree with the Examiner that Zhang determines ellipses, center points of the ellipses and other parameters to describe the shape of the silhouette (Ans. 12; Zhang [0022], [0023], [0026], and [0027]). Therefore, Zhang describes that the disclosed ellipse fitting technique (keypoint selection technique) is applied only to pixels belonging to the extracted object. Accordingly, the Examiner did not err in finding the combination of Karmann and Zhang teaches or suggests “selecting keypoints using a keypoint selection technique which is applied only to pixels belonging to the extracted object” as recited in independent claims 2 and 6. Dependent claims 3, 4, 7, and 8 were not separately argued and thus fall with their respective independent claims. Therefore, we sustain the rejection of claims 2-4 and 6-8 under 35 U.S.C. § 103(a) for obviousness over Karmann and Zhang. Appeal 2010-007698 Application 11/821,767 7 ISSUE 2 35 U.S.C. § 103(a): claims 11, 12, 15, and 16 Appellants assert their invention is not obvious over Karmann, Zhang, and Courtney because Courtney does not teach “constructing a temporal graph where … each said link forms a weighted edge between linked nodes” as recited in claims 11 and 15 (App. Br. 9-11). Specifically, Appellants contend Courtney teaches “primary links are solid lines and secondary links are dashed” in FIG. 9 but does not teach the links being weighted edges (App. Br. 10). Issue 2: Has the Examiner erred in concluding the combination of Karmann, Zhang, and Courtney teaches or suggests “constructing a temporal graph where each segment forms a node in the graph and each said link forms a weighted edge between linked nodes” as recited in independent claims 11 and 15? ANALYSIS We agree with the Examiner that Courtney teaches the links are weighted edges (Ans. 15-17). Initially we find Appellants have not explicitly defined “weighted edge” in their Specification. Although Appellants contend their Specification describes “weighted edges,” these portions of the Specification do not give rise to a definition of “weighted edges” (App. Br. 10). Instead, these portions of Appellants’ Specification provide various embodiments (App. Br. 10). Appellants additionally submit “that one of ordinary skill in the pertinent art would know that a weighted Appeal 2010-007698 Application 11/821,767 8 edge means an edge that is weighted by a numerical factor” (Reply Br. 3). However, Appellants have not provided sufficient evidence or argument to persuade us of error in the Examiner’s interpretation. We agree with the Examiner that Courtney describes a graph linked by solid and dashed lines and that Figure 9 discloses primary and secondary links (Ans. 16-17; col. 8, ll. 56- 63; col. 9, ll. 10-12; Fig. 9). Courtney discloses solid lines representing primary links and the secondary links are dashed (col. 3, ll. 35-40). In Courtney, pairs of mutual nearest neighbors are connected using a “primary” link to indicate that they are highly likely to represent the same real-world object in successive video frames, i.e., are most reliably tracked (col. 8, ll. 56-59; see also col. 9, ll. 11-12). Video objects (V-objects) that are tracked with less confidence but display evidence that they might result from the same real-world object are joined by “secondary” links (col. 8, ll. 60-63). We agree with the Examiner that the links are weighted by the confidence level – highly likely or less likely (Ans. 15-16). Thus, we find each of these links form a weighted edge between linked nodes. Accordingly, the Examiner did not err in finding Courtney teaches “constructing a temporal graph where each segment forms a node in the graph and each said link forms a weighted edge between linked nodes” as recited in independent claims 11 and 15. Dependent claims 12 and 16 were not separately argued and thus fall with their respective independent claims. Therefore, we sustain the rejection of claims 11, 12, 15, and 16 under 35 U.S.C. § 103(a) for obviousness over Karmann, Zhang, and Courtney. Appeal 2010-007698 Application 11/821,767 9 DECISION The Examiner’s rejection of claims 2-4 and 6-8 under 35 U.S.C. § 103(a) as being unpatentable over Karmann and Zhang is affirmed. The Examiner’s rejection of claims 11, 12, 15, and 16 under 35 U.S.C. § 103(a) as being unpatentable over Karmann, Zhang, and Courtney is affirmed. No time period for taking any subsequent action in connection with this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv) (2011). AFFIRMED tj Copy with citationCopy as parenthetical citation