IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`
`: 90/012,876
`Control No.
`: 7,932,923
`Patent No.
`: May 23, 2013
`Filed
`Customer No. : 06449
`
`Art Unit
`Examiner
`Conf.No.
`Atty. No.
`
`: 3992
`: Adam L. Basehoar
`: 6419
`: 4079-116
`
`Title: VIDEO SURVEILLANCE SYSTEM EMPLOYING VIDEO PRIMITNES
`37 C.F.R. § 1.132 DECLARATION OF KENNETH A. ZEGER
`I, Kenneth A. Zeger, declare as follows:
`My name is Kenneth A. Zeger. I am a Full Professor of Electrical and Computer
`1.
`Engineering at the University of California, San Diego (UCSD). I understand that my
`declaration is being submitt~d in connection with the above-referenced reexamination
`proceeding pending in the United States Patent and Trademark Office.
`Qualifications, Background, and Experience
`I.
`I have studied, taught, and practiced electrical and computer engineering for over
`2.
`
`thirty years.
`
`I attended the Massachusetts Institute of Technology ("MIT") and earned
`3.
`Bachelors (SB) and Masters (SM) of Science Degrees in Electrical Engineering and Computer
`Science in 1984. I earned a Masters of Arts (MA) Degree in Mathematics in 1989 from the
`University of California, Santa Barbara. I also earned my Ph.D. in Electrical and Computer
`Engineering from the University of California, Santa Barbara in 1990.
`I have held the position of Full Professor of Electrical and Computer Engineering
`4.
`at UCSD since 1998, having been promoted from Associate Professor after two years at UCSD.
`'
`I teach courses full-time at UCSD in the fields of Electrical and Computer Engineering, and
`specifically in subfields including information theory and image coding, at the undergraduate
`and graduate levels. Prior to my employment at UCSD, I taught and conducted research as a
`faculty member at the University of Illinois, Urbana-Champaign for four years, and at the
`University of Hawaii for two years.
`I am president of Zunda LLC ("Zunda") a California company· located in San
`5.
`Diego, California. Zunda provides expert witness and technical consulting services in the fields
`of electrical engineering and computer hardware/software.
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page2
`
`My twenty-plus years of industry experience includes consulting work for the
`6.
`United States Department of Defense as well as for private companies such as Xerox, Nokia,
`MITRE, ADP, and Hewlett-Packard. The topics upon which I provide consulting expertise
`include image, video, and speech coding; ·data compression; networks; digital communications;
`pattern recognition; computer software; and mathematical analyses.
`I have authored almost 70 peer-reviewed journal articles, the majority of which
`7.
`are on the topic of compression or information theory. I have also authored over 100 papers at
`various conferences and symposia over the past twenty-plus years, such as the IEEE
`International Symposium on Information Theory, the International Conference on Image
`Processing, and the Data Compression Conference.
`I was elected a Fellow of the IEEE in 2000, an honor bestowed upon only a small
`8.
`percentage of IEEE members. I was awarded the National Science Foundation Presidential
`· Young Investigator Award in 1991, which included $500,000 in research funding. I received
`this award one year after receiving my Ph.D.
`I have served as an Associate Editor for the IEEE Transactions on Information
`9.
`Theory publication and have been an elected member of the IEEE Information Theory Board of
`Governors for three, three-year terms. I organized and have been on the technical advisory
`committees of numerous workshops and symposia in the areas of image coding, information
`theory, and data compression. I regularly review submitted journal manuscripts, government
`funding requests, conference proposals, student theses, and textbook proposals. I also have
`given many lectures at conferences, universities, and companies on topics in image coding, data
`compression, and information theory.
`I have extensive experience in electronics hardware and computer software, from
`10.
`academic studies, work experience, and supervising students. I personally program computers
`on an almost daily basis and have fluency in many different computer languages.
`A more complete recitation of my professional experience including a list of my
`11.
`publications is set forth in my curriculum vitae, attached to my declaration as Exhibit Zl.
`Compensation and Engagement
`II.
`Zunda is being compensated for my work in this matter by Rothwell, Figg, Ernst
`12.
`& Manbeck, at my current rate of $690 per hour. Neither Zunda nor I have any personal or
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page3
`
`financial stake or interest in the outcome of the above-referenced reexamination or any related
`litigation matter. Neither Zunda's nor my compensation is dependent upon my testimony or the
`outcome of this proceeding or any related litigation matter. Neither Zunda nor I have any
`relationship with or financial interest in the assignee of U.S. Patent No. 7,932,923 ("the '923
`patent"), ObjectVideo, Inc.
`III. The Reexamination Proceeding
`It is my understanding that, on May 23, 2013, an anonymous Requestor ("the
`13.
`Requester") filed a Request for Ex Parte Reexamination (the "Request") with the United States
`Patent and Trademark Office (the "Office") requesting reexamination of the '923 patent and that,
`on June 17, 2013, the Office issued an Order granting the Request (the "Order"). I understand
`that the Office determined that the Request established a substantial new question of
`patentability with respect to claims 1-41 of the '923 patent. Thus, it is my understanding that the
`Office is reexamining claims 1-41 of the '923 patent.
`It is also my understanding that, on August 30, 2013, the Office issued an Office
`14.
`Action (the "Office Action" or "OA"). In the Office Action, claims 1-41 of the '923 patent are
`rejected under 35 U.S.C. § 102(b) as anticipated by certain of the references identified in the
`Office Action and/or under 35 U.S.C. § 103(a) as being obvious in view of certain of the
`references identified in the Office Action.
`I have read and understand the '923 patent, its prosecution history, and the
`15.
`references cited in the '923 patent. I have read and understand the Request, the Order, the Office
`Action, and the references cited in the Office Action. I have also read and understand the
`comments filed by Bosch, the third party requester, on July 11, 2012, in the terminated inter
`partes reexamination (Control No. 95/001,912) of U.S. Patent No. 7,868,912 ("Bosch's
`comments").
`
`16.
`
`I was asked to consider and address the following rejections of claims 1-41 of the
`'923 patent raised in the Office Action:
`(i) Claims 1-9, 13, 15-18, 20-30, 34, and 36-40 under 35 U.S.C. § 102(b) as anticipated
`by "Object Oriented Conceptual Modeling of Video Data," Day et al. ("Day-f');
`(ii) Claims 1-7, 9-13, and 15-28 under 35 U.S.C. § 102(b) as anticipated by U.S. Patent
`No. 5,969,755 to Courtney ("Courtney '755");
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page4
`
`(iii) Claims 1-7, 9-13, and 15-28 under 35 U.S.C. § 102(b) as anticipated by "Object
`Tracking and Event Recognition in Biological Microscopy Videos," Shotton et al.
`("Shotton");
`(iv) Claims 14 and 35 under 35 U.S.C. § 103(a) as unpatentable over Day-I;
`(v) Claims 10, 19, 31, and 41 under 35 U.S.C. § 103(a) as unpatentable over Day-I in
`view of U.S. Patent No. 6,628,835 to Brill et al. ("Brill");
`(vi) Claims 11, 12, 32, and 33 under 35 U.S.C. § 103(a) as unpatentable over Day-I in
`view of "Spatio-Temporal Modeling of Video Data for On-Line Object Oriented
`Query Processing," Day et al. ("Day-II'');
`(vii) Claim 14 under 35 U.S.C. § 103(a) as unpatentable over Courtney '755;
`(viii) Claim 14 under 35 U.S.C. § 103(a) as unpatentable over Shotton;
`(ix) Claims 8 and 29-41 under 35 U.S.C. § 103(a) as unpatentable over Shotton in view
`of Brill; and
`(x) Claims 1-41 under 35 U.S.C. § 103(a) as unpatentable over European Patent
`Application Publication No. EP 0967584A2 to Courtney ("Courtney '584") in
`view of Brill.
`My opinions regarding the rejections are set forth below.
`
`IV. Applicable Laws/Rule
`Claim Interpretation
`A.
`I understand that, during reexamination, the pending claims must be given their
`17.
`broadest reasonable interpretation consistent with the specification, and that the broadest
`reasonable interpretation of the claims must also be consistent with the interpretation that those
`skilled in the art would reach.
`B.
`Priority
`.
`I understand that claims of an application that is a continuation or continuation-in-
`18.
`part of an earlier U.S. application or international application which are fully supported under 35
`U.S.C. § 112 by the earlier parent application have the effective filing date of that earlier parent
`application. A claim is adequately disclosed/fully supported under 35 U.S.C. § 112 by an earlier
`parent application if the earlier parent application satisfies the written description requirement.
`To satisfy the written description requirement, a patent specification must describe the claimed
`invention in sufficient detail that one skilled in the art can reasonably conclude that the inventor
`had possession of the claimed invention.
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page5
`
`Anticipation (35 U.S.C. § 102)
`C.
`To support a rejection based on 35 U.S.C. § 102, I understand that the Examiner
`19.
`bears the burden of showing that a single prior art reference discloses all of the elements of the
`claim, arranged in the same manner as required by the claim, either explicitly or inherently.
`Obviousness (35 U.S.C. § 103)
`D.
`I also understand that a claim is not patentable if the differences between the
`20.
`subject matter of the claim and the disclosure of the prior art are such that ~e subject matter of
`the claim, as a whole, would have been obvious at the time of invention to a person having
`ordinary skill in the art to which the subject matter pertains.
`In determining obviousness, I understand that it is necessary to consider the scope
`21.
`. and content of the prior art; the differences between the prior art and the claims at issue; the level
`of ordinary skill in the pertinent art; and any objective evidence of non-obviousness related to the
`alleged merits of the claimed invention (which I understand is referred to as "objective indicia of
`non-obviousness"), such as commercial success, long-felt but unsolved needs, industry
`recognition, failure of others, and copying.
`In determining obviousness based on a combination of prior art references, I also
`22.
`understand that evidence of some reason to combine the teachings is required to make the
`combination, and thus such evidence must be considered, along with any evidence that one or
`more of the references would have taught away from the claimed invention at the time of the
`invention.
`I have been informed that the hypothetical person of ordinary skill in the art is
`23.
`presumed to know all of the teachings known in the art at the time the alleged invention was
`made. That person is presumed to have the technical competence and experience of skilled
`artisans working in the area of the subject invention and of the manner in which problems were
`solved. Factors that may be considered in determining the level of ordinary skill in the art
`include the types of problems encountered in the art, prior art solutions to those problems, the
`rapidity with which innovations are made, the sophistication of the technology, and the
`educational level of active workers in the field.
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page6
`
`V.
`
`Priority
`Patent Claims 1-41
`A.
`I have compared U.S. Patent Application No. 09/987,707 (the '"707 application"),
`24.
`filed on November 15, 2001, with the specification of the '923 patent and conclude that they are
`identical ( except for the cross-reference to related applications, clarifications of the references
`describing moving target detection, and claims in the '923 patent). Thus, the '707 application
`fully supports claims 1-41 in the '923 patent. Below, citations to the specification of the '923
`patent refer to the '707 application, which is incorporated by reference in the '923 patent. '923
`patent at col. 1, lines 7-11.
`New Claims 42-81
`B.
`I have examined, in detail, new claims 42-81 of Patent Owner's amendment and
`25.
`reply, the '923 patent, and the '707 application, filed on November 15, 2001. It is my opinion
`that both the '923 patent and the '707 application fully support new claims 42-81 of Patent
`Owner's amendment and reply.
`New claim 42 is similar to patent claim 20 but additionally requires "applying the
`26.
`new user rule to the combination of the attributes to detect the event." Support for this feature
`can be found throughout the specification of the '923 patent. See, e.g., '707 application at en 66
`("An operator is provided with maximwn flexibility in configuring the system by using event
`discriminators. Event discriminators are identified with one or more objects (whose descriptions
`are based on video primitives), along with one or more optional spatial attributes, and/or one or
`more optional temporal attributes."); 'l[ 80 ("An event discriminator is described in terms of video
`· primitives."); CJI 118 ("In block 44, event occurrences are extracted from the video primitives
`using event discriminators .... The event discriminators are used to filter the video primitives to
`determine if any event occurrences occurred .... The event discriminator checks all video
`primitives being generated according to FIG. 5 and determines if any video primitives exist
`which have the following properties: .... ").
`New claims 43-46 are similar to patent claims 1, 9, 20, and 22, respectively, but
`27.
`explicitly require "filtering." The filtering requirement is fully supported by the specification of
`the '923 patent, which discloses that "[t]he event discriminators are used to filter the video
`primitives to determine if any event occurrences occurred." '707 application at <JI 118 (emphasis
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page?
`
`added). See also id. at 'I[ 30 ("A need exists to filter video surveillance data to identify desired
`portions of the video surveillance data." (emphasis added)) & 132 ("An object of the invention
`is to filter video surveillance data to identify desired portions of the video surveillance data."
`(emphasis added)).
`New claims 47-50 are similar to patent claims 1, 9, 20, and 22, respectively, but
`28.
`additionally require attribute detection "in real time." The "real time" attribute detection is fully
`supported by the specification of the '923 patent, which discloses, for example, that "video
`_ primitives are extracted in real time from the source video." '707 application at 'I[ 118. See also
`id. at CWl[ 33, 65, 104.
`New claims 51-54 are similar to patent claims 1, 9, 20, and 22, respectively, but
`29.
`additionally require a "velocity" attribute. The "velocity" attribute detection is fully supported
`by the specification of the '923 patent. See, e.g., '707 application at 180 (disclosing "a velocity"
`as an example of a video primitive that "refers to an observable attribute of an object viewed in a
`video feed") & 188 (disclosing "a velocity (e.g., speed and ~irection) in image space" and "an
`approximate velocity (e.g., speed and direction) in a three-dimensional representation of the
`environment" as examples of a feature of a salient motion video primitive).
`New claims 55-58 are similar to patent claims 1, 9, 20, and 22, respectively, but
`30.
`additionally require that applying the new user rule to the plurality of detected attributes
`comprises applying the new user rule to "only" the plurality of detected attributes. The user rule
`application to "only" the detected attributes feature is fully supported by the '923 patent, which
`discloses that "[t]he video content can be reanalyzed with the additional embodiment in a
`relatively short time because only the video primitives are reviewed and because the video source
`is not reprocessed." '707 application at <JI 148 (emphasis added). See also id. at «JI 67 ("By
`analyzing previously processed video, the system can perform inference analysis based on
`previously recorded video primitives, which greatly improves the analysis speed of the computer
`system."); <JI 68 ("the system of the invention uses event discriminators based on video primitives
`as the primary tasking mechanism"); 1 118 ("In block 44, event occurrences are extracted from
`the video primitives using event discriminators. The video primitives are determined in block 42,
`and the event discriminators are determined from tasking the system in block 23. The event
`discriminators are used to filter the video primitives to determine if any event occurrences
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page8
`
`occurred."); 'lI 148 ("the system analyses [sic] archived video primitives with event
`discriminators to generate additional reports, for example, without needing to review the entire
`source video"); 'II 148 (Reviewing only the video primitives "provides a great efficiency
`improvement over current state-of-the-art systems because processing video imagery data is
`extremely computationally expensive, whereas analyzing the small-sized video primitives
`abstracted from the video is extremely computationally cheap.").
`New claims 59-61 are similar to patent claims 8, 29, and 30, respectively, but
`31.
`additionally require that the identified event be "the first and second objects coming together."
`The identified event that is "the first and second objects coming together" is fully supported by
`the specification of the '923 patent, which discloses "two objects come together" as an example
`of "an event discriminator for multiple objects." '707 application at 'JI 99.
`New claims 62-64 are similar to patent claims 8, 29, and 30, respectively, but
`32.
`explicitly require "filtering" and additionally require that the identified event be "the first and
`second objects coming together." The identified event that is "the first and second objects
`coming together" is fully supported by the specification of the '923 patent, which discloses "two
`objects come together" as an example of "an event discriminator for multiple objects." '707
`application at 'lI 99. The filtering requirement is also fully supported by the specification of the
`'923 patent, which discloses that "[ t ]he event discriminators are used to filter the video
`primitives to determine if any event occurrences occurred." '707 application at 'l[ 118 ( emphasis
`added). See also id. at 'lI 30 ("A need exists to filter video surveillance data to identify desired
`portions of the video surveillance data." (emphasis added)) & 'l[ 32 ("An object of the invention
`is to filter video surveillance data to identify desired portions of the video surveillance data."
`(emphasis added)).
`New claims 65-68 are similar to patent claims 1, 9, 20, and 22, respectively, but
`33.
`additionally require that "none of the detected attributes refers to the object engaged in an
`activity." This additional feature is fully supported by the specification of the '923 patent, which
`discloses the identification of events solely from attributes that are non-event characteristics of
`an object (i.e., non-event attributes). See, e.g., '707 application at cn'iI 100 & 118. For example,
`the '923 patent discloses, as an example, that "an event discriminator can be looking for a
`'wrong way' event as defined by a person traveling the 'wrong way' into an area between 9:00
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page9
`
`a.m. and 5:00 p.m. The event discriminator checks all video primitives being generated
`according to FIG. 5 and determines if any video primitives exist which have the following
`properties: a timestamp between.9:00 a.m. and 5:00 p.m., a classification of 'person' or 'group of
`people', a position inside the area, and a 'wrong' direction of motion." '707 application at <JI 118.
`The attributes used to determine if the wrong way event occurred are times, object types,
`positions, and directions, and all of these attributes are non-event attributes. In other words,
`none of these attributes used to determine if the wrong way event occurred refers to an object
`engaged in activities. Another example in the specification of the '923 patent of an event
`identified from solely non-event attributes is the "object crosses a line" event, which is disclosed
`as an example of an event discriminator for an object and a spatial attribute." '707 application at
`<J[ 100. A skilled person would understand from reading the specification of the '923 patent that
`detecting an object crossing a line would be accomplished by detecting purely non-event
`attributes based on time and location, namely where the object is located as a function of time
`and where the line is located.
`New claims 69-72 are similar to patent claims 1, 9, 20, and 22, respectively, but
`34.
`explicitly require "filtering" and additionally require that "none of the detected attributes refers
`to the object engaged in an activity." The specification of the '923 patent fully supports "none of
`the detected attributes refers to the object engaged in an activity" and "filtering" for the reasons
`explained above in paragraphs 27 & 33.
`New claims 73 and 74 are similar to patent claims 1 and 22, respectively, but
`35.
`explicitly require "filtering" and additionally require attribute detection "in real time." The
`specification of the '923 patent fully supports "filtering" and "real time" attribute detection for
`the reasons explained above in <Jr![ 27 & 28.
`New claim 75 is similar to patent claim 1 but additionally requires attribute
`36.
`detection "in real time" and a "velocity" attribute. New claim 76 is similar to patent claim 1 but
`explicitly requires "filtering" and additionally requires a "velocity" attribute. New claim 77 is
`similar to patent claim 1 but explicitly requires "filtering" and additionally requires attribute
`detection "in real time" and a "velocity" attribute. The specification of the '923 patent fully
`supports each of the "filtering," "real time" attribute detection, and "velocity" attribute for the
`reasons explained above in <Jr![ 27-29.
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page 10
`
`New dependent claims 78-81 depend on new claims 43-46, respectively, and
`37.
`additionally require that "the filtering comprises filtering the plurality of detected attributes to
`determine if the event occurred" (claims 78, 79, and 81) or that "the filtering comprises filtering
`the combination of attributes to determine if the event occurred" ( claim 80). The attribute
`filtering requirement is fully supported by the specifi9ation of the '923 patent, which discloses
`that "[t]he event discriminators are used to filter the video primitives to determine if any event
`occurrences occurred." '707 application at 'II 118 (emphasis added). See also id. at 'II 30 ("A
`need exists to filter video surveillance data to identify desired portions of the video surveillance
`data." (emphasis added)) & 132 ("An object of the invention is to filter video surveillance data
`to identify desired portions of the video surveillance data." (emphasis added)).
`Claim Construction
`VI.
`Attributes and Events
`A.
`Given its broadest reasonable interpretation, it is my opinion that the claim
`38.
`language "attributes," as used in claims 1-41 of the '923 patent, means: observable
`characteristics. My "observable characteristics" definition is informed by how this phrase is
`used in the specification, including the claims, of the '923 patent. See '707 application at 1 80.
`Further, the plain and ordinary meaning of the term "attributes" is that they are observable
`characteristics of something. See Exhibit Z2, Webster's New World Dictionary of Computer
`Terms (5th Ed. 1994) (providing "[a] characteristic quality of a data type, data structure, element
`of a data model, or system" as a definition of "attribute"). The specification of the '923 patent
`supports this plain meaning, such as when it states that:
`A video primitive refers to an observable attribute of an object viewed in a video
`feed. Examples of video primitives include the following: a classification; a size;
`a shape; a color; a texture; a position; a velocity; a speed; an internal motion; a
`motion; a salient motion; a feature of a salient motion; a scene change; a feature
`of a scene change; and a pre-defined model.
`'707 application at 'II 30.
`Given its broadest reasonable interpretation, it is my opinion that the claim
`39.
`language "event" means: one or more objects engaged in an activity. The specification expressly
`defines an "event" as "one or more objects engaged in an activity." See '707 application at <JI 48
`("An 'event' refers to one or more objects engaged in an activity."). My "one or more objects
`engaged in an activity" definition is informed by this express definition in the specification.
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page 11
`
`The specification of the '923 patent refers to the claimed "attributes" as
`40.
`"primitives,", and gives numerous examples of attributes/primitives, such as: "a classification; a
`size; a shape; a color; a texture; a position; a velocity; a speed; an internal motion; a motion; a
`salient motion; a feature of a salient motion; a scene change; a feature of a scene change; and a
`pre-defined model." '707 application at <J[ 80. The '923 patent also gives numerous examples of
`events, such as: an object appears; a person appears; a red object moves faster than 10 mis, two
`objects come together; a person exits a vehicle; a red object moves next to a blue object, an
`object crosses a line; an object enters an area; a person crosses a line from the left, an object
`appears at 10:00 p.m.; a person travels faster then [sic] 2 mis between 9:00 a.m. and 5:00 p.m.; a
`vehicle appears on the weekend, a person crosses a line between midnight and 6:00 a.m.; a
`vehicle stops in an area for longer than 10 minutes; a person enters an ·area between midnight and
`6:00 a.m.; and a security service is notified. Id. at 'Ir)[ 98-103. Generally speaking, attributes are
`simpler concepts than events.
`41. More precisely, the '923 patent teaches that multiple detected attributes are to be
`examined and then, based upon such attributes, a decision is made as to whether or not certain
`events occurred. '707 application at 'I[ 118. The system described in the '923 patent is
`configured to detect attributes by analyzing a video, but the choice of which attributes the system
`is configured to detect is not based upon which events are later to be identified. See id. at 'l[ 79.
`In fact, tasking of the system to identify one or more events from the detected attributes is not
`even necessary. Id. at 'I[ 79 ("Tasking occurs after calibration in block 22 and is optional.
`Tasking the video surveillance system involves specifying one or more event discriminators.
`Without tasking, the video surveillance system operates by detecting and archiving video
`primitives and associated video imagery without taking any action, as in block 45 in FIG. 4."). 1
`While the specification of the '923 patent does not explicitly use the term "independence"
`(outside of the claims themselves), a person of ordinary skill in the art of the '923 patent would
`understand the '923 patent to teach that the choice of which events are to be identified (i.e.,
`tasking) is made at a time after configuration of the system to detect attributes, and furthermore
`
`1 Tasking the system determines the event discriminators (i.e., user rules) that may be selected
`for use in identifying events. See '707 application at <JI 118 ("event discriminators are determined
`from tasking the system in block 23" and "are used to filter the video primitives to determine if
`any event occurrences occurred").
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page 12
`
`that the choice of which attributes the system is configured to detect is not dictated/determined
`by which events the system might later be tasked to identify. See id. at')[ 79. This indeed means
`that the attributes collected are independent of the events identified. Moreover, this
`. independence of the attributes from the events in the '923 patent means that that the selection of
`which attributes to detect is not based upon a predefined list of events to be determined.
`The specification of the '923 patent does not prohibit an identified event from
`42.
`being the same as a detected attribute, but such a scenario is not within the scope of the claims of
`the '923 patent, which require that the identified event not be one of the detected attributes.
`Claims 1, 9, and 22 ("identifying an event of the object that is not one of the detected attributes
`of the object"); claims 8 and 29 ("identifying an event that is not one of the detected attributes of
`the first and second objects"); claim 20 ("detect an event that is not one of the detected
`attributes"); claim 30 ("the event not being one of the detected attributes").
`To help clarify the distinction between the claimed "events" and the claimed
`43.
`"attributes," I will illustrate the concepts with an example based upon events such as an object
`moving, entering, or growing. The specification of the '923 patent discloses that "an 'event'
`refers to one or more objects engaged in an activity" and provides, as examples of an activity,
`"entering; exiting; stopping; moving; raising; lowering; growing; and shrinking." '707
`application at 'ffi[ 46 & 48. Multitudes of attributes/characteristics of an object that is engaging in
`any one such activity may be associated with the event that refers to the activity. The
`attributes/characteristics may include, for example, one or more of the object's position, width,
`length, (linear) speed, velocity, acceleration, third order and higher derivatives of motion vs.
`time, direction of motion, momentum, rotation, angular velocity, moment of inertia, angular
`momentum, occlusions, shading, proximity to nearby objects, etc. None of these
`attributes/characteristics is itself an "activity." Instead, these attributes/characteristics are
`numerical descriptions of particular observable aspects of the object that may be engaging in an
`activity. One might be able to logically deduce, from one or more of these
`attributes/characteristics, a particular activity in which an object is engaging, but that involves
`inference. Each of these example characteristics is an attribute that does not refer to an activity
`in which the object is engaged. In contrast, a hypothetical Boolean variable such as ls_Entering,
`Is_Exiting, ls_Moving, etc., which is true if the object is engaging in the associated activity and
`
`

`

`Declaration of Kenneth A. Zeger
`Control No. 90/012,876
`Page 13
`
`false if the object is not engaging in such activity, would be an example of an attribute that refers
`to an activity in which the object is engaged, because, in this case, no deductive reasoning is
`required to determine that a particular event referring to the object engaged in the activity
`
`occurred.
`
`Inferring events based on the attributes (or primitives) provides several
`44.
`advantages. First, "[a]n operator is provided with maximum flexibility in configuring the system
`by using event discriminators," '707 application at <J[ 66, as opposed to the limitations associated
`with using prerecorded events. Second, inference analysis based on previously extracted
`attributes "greatly improves the analysis speed of the computer system" as the system can
`process only the attributes instead of reprocessing the video. See id. at<][ 67. Third, inventing a
`system for analyzing "small-sized video primitives abstracted from the video" has many
`corresponding size-based benefits. Id. at <J[ 148. For example, the storage space and bandwidth
`necessary to manage the small-sized attributes is far less than for managing the video itself, even
`if the video is highly compressed. See id.
`User Rule Filtering Claim Elements
`Claims 1-41 of the '923 patent all contain limitations that require user rule
`45.
`filtering of detected attributes. The following claim elements (the "user rule filtering elements")
`
`B.
`
`incorporat