`United States Patent
`(10) Number:
`US 7,932,923 Cl
`
`(45) Certificate Issued:Liptonetal. May 21, 2014
`
`
`US007932923C1
`
`(54) VIDEO SURVEILLANCE SYSTEM
`EMPLOYING VIDEO PRIMITIVES
`
`continuation-in-part of application No. 09/694,712,
`filed on Oct. 24, 2000, now Pat. No. 6,954,498.
`
`(75)
`
`Inventors: Alan J. Lipton, Falls Church, VA (US);
`Thomas M.Strat, Pakton, VA (US);
`Péter L. Venetlaner, McLean, VA (US);
`MarkC. Allmen, Morrison, CO (US);
`William E. Severson, Littleton, CO
`(US); Niels Haering, Arlington, VA
`(US); Andrew J. Chosak, McLean, VA
`(US); Zhong Zhang, Herndon, VA (US);
`Matthew F. Frazier, Arlington, VA
`(US); James S. Seekas, Arlington, VA
`(US); Tasuki Hirata, Silver Spring, MD
`(US); John Clark, Leesburg, VA (US)
`
`(73) Assignee: Diamondback Vision, Inc., Reston, VA
`(US)
`
`Reexamination Request:
`No. 90/012,876, May 23, 2013
`
`Reexamination Certificate for:
`Patent No.:
`7,932,923
`Issued:
`Apr. 26, 2011
`Appl. No.:
`12/569,116
`Filed:
`Sep. 29, 2009
`
`Related U.S. Application Data
`
`(63) Continuation of application No. 09/987,707, filed on
`Nov.
`15,
`2001, now abandoned, which is
`a
`
`(51)
`
`Int. Cl.
`HOAN 7/18
`
`(2006.01)
`
`(52) U.S. CL
`USPC oie ccesceceesesseseeseeeecseneeseteeees 348/143
`
`(58) Field of Classification Search
`None
`
`See application file for complete search history.
`
`(56)
`
`References Cited
`
`To view the complete listing of prior art documents cited
`during the proceeding for Reexamination Control Number
`90/012,876, please refer to the USPTO’s public Patent
`Application Information Retrieval (PAIR) system under the
`Display References tab.
`
`Primary Examiner — Adam L Basehoar
`
`(57)
`
`ABSTRACT
`
`A video surveillance system is set up, calibrated, tasked, and
`operated. The system extracts video primitives and extracts
`event occurrences from the video primitives using event dis-
`criminators. The system can undertake a response, such as an
`alarm, based on extracted event occurrences.
`
`45
`
`NOUS
`
`
`
`
`
`
`
`undertake
`archive
`extract
`obtain
`extract
`
`
`
`
`response,
`video
`source
`video
`event
`
`
`
`
`
`as appropriate
`primitives
`video
`primitives
`occurrences
`
`
`
`
`
`
`
`44
`
`Canon Ex. 1020 Page 1 of 3
`
`Canon Ex. 1020 Page 1 of 3
`
`
`
`US 7,932,923 Cl
`
`1
`EX PARTE
`
`REEXAMINATION CERTIFICATE
`
`ISSUED UNDER35 U.S.C. 307
`
`THE PATENT IS HEREBY AMENDED AS
`INDICATED BELOW.
`
`Matter enclosed in heavy brackets [ ] appeared in the
`patent, but has been deleted andis no longer a part of the
`patent; matter printedin italics indicates additions made
`to the patent.
`
`ONLY THOSE PARAGRAPHSOF THE
`SPECIFICATION AFFECTED BY AMENDMENT
`ARE PRINTED HEREIN.
`
`Column 1, lines 8-12:
`
`This application claimsthe priority to and is a continuation
`ofUS. patent application Ser. No. 09/987,707, filed Nov. 15,
`2001 now abandoned, which claimspriority to and is a con-
`tinuation-in-part ofU.S. patent application Ser. No. 09/694,
`712,filed on Oct. 24, 2000, now U.S. Pat. No. 6,954,498, each
`of which is incorporated herein by referencein their entirety.
`
`AS A RESULT OF REEXAMINATION, IT HAS BEEN
`DETERMINEDTHAT:
`
`Claims 1, 8, 9, 20, 22, 29 and 30 are determined to be
`patentable as amended.
`Claims 2-7, 10-19, 21, 23-28 and 31-41, dependent on an
`amended claim, are determinedto be patentable.
`1. A method comprising:
`detecting an object in a video from a single camera;
`detecting a plurality ofattributes of the object by analyzing
`the video from said single camera,
`the plurality of
`attributes including at least one of a physical attribute
`and a temporal attribute, each attribute representing a
`characteristic of the detected object;
`selecting a new user rule after detecting the plurality of
`attributes; and
`after detecting the plurality ofattributes and after selecting
`the new userrule, identifying an event of the objectthat
`is not one of the detected attributes of the object by
`applying the new userrule to the plurality of detected
`attributes, wherein the applying the new userrule to the
`plurality of detected attributes comprises applying the
`new user rule to only the plurality ofdetected attributes;
`wherein the plurality of attributes that are detected are
`independent of which eventis identified,
`wherein the step of identifying the event of the object
`identifies the event without reprocessing the video, and
`wherein the event of the object refers to the object engaged
`in an activity.
`8. A method comprising:
`detecting first and second objects in a video from a single
`camera;
`detecting a plurality of attributes of each of the detected
`first and second objects by analyzing the video from said
`single camera, eachattribute representing a characteris-
`tic of the respective detected object;
`selecting a new user rule; and
`after detecting the plurality of attributes, identifying an
`eventthatis not one of the detected attributesofthefirst
`and second objects by applying the new userrule to the
`plurality of detected attributes, wherein the applying the
`
`2
`new user rule to theplurality ofdetected attributes com-
`prises applying the new userrule to only the plurality of
`detected attributes;
`wherein the plurality of attributes that are detected are
`independent of which eventis identified,
`wherein the step of identifying an event of the object com-
`prises identifying a first event of the first object interact-
`ing with the second object by analyzing the detected
`attributes of the first and second objects,the first event
`not being one ofthe detected attributes, and
`wherein the event ofthe object refers to the object engaged
`in an activity.
`9. A video device comprising:
`means for detecting an object in a video from a single
`camera;
`meansfor detecting a plurality of attributes ofthe object by
`analyzing the video from said single camera,the plural-
`ity of attributes includingat least a physicalattribute and
`a temporal attribute, each attribute representing a char-
`acteristic of the detected object;
`a memory storing the plurality of detected attributes;
`means for selecting a new userrule after the plurality of
`detected attributes are stored in memory; and
`meansfor identifying an event of the object that is not one
`of the detected attributes of the object by applying a
`selected new userruleto the plurality ofattributes stored
`in memory, for identifying the event independent of
`whenthe attributes are stored in memory and for iden-
`tifying the event without
`reprocessing the video,
`wherein the applying the new userrule to theplurality of
`detected attributes comprises applying the new userrule
`to only the plurality ofdetected attributes, and
`wherein the event ofthe object refers to the object engaged
`in an activity.
`20. A method comprising:
`providing a video device which detects an object upon
`analyzing a video from a single camera and which
`detects plural attributes of the detected object upon ana-
`lyzing the video from said single camera, the plurality of
`attributes including at least a physical attribute and a
`temporalattribute; and
`then,selecting a rule, which is not a rule used to detect any
`individualattribute, as a new userrule, the new user rule
`providing an analysis of a combination ofthe attributes
`to detect an event
`that
`is not one of the detected
`
`5
`
`30
`
`40
`
`45
`
`attributes, wherein the analysis ofthe combination ofthe
`attributes to detect the event comprises analyzing only
`the combination ofthe attributes,
`wherein theattributes to be detected are independentofthe
`event to be detected, and
`wherein the event ofthe object refers to the object engaged
`in an activity.
`22. A non-transitory computer-readable storage medium
`containing instructions that when executed by a computer
`55 system cause said computer system to implementthe follow-
`ing method comprising:
`detecting an object in a video from a single camera;
`detecting a plurality of attributes ofthe object by analyzing
`the video from said single camera,
`the plurality of
`attributes including at least one of a physical attribute
`and a temporal attribute, each attribute representing a
`characteristic of the detected object;
`selecting a new user rule after detecting the plurality of
`attributes; and
`after detecting theplurality of attributes andafter selecting
`the new userrule, identifying an event of the object that
`is not one of the detected attributes of the object by
`
`60
`
`65
`
`Canon Ex. 1020 Page 2 of 3
`
`Canon Ex. 1020 Page 2 of 3
`
`
`
`US 7,932,923 Cl
`
`3
`applying the new userrule to the plurality of detected
`attributes, the event of the object being identified with-
`out reprocessing the video, wherein the applying the new
`user rule to theplurality ofdetected attributes comprises
`applying the new user rule to only the plurality of 5
`detected attributes;
`wherein the plurality of attributes that are detected are
`independent of which event is identified, and
`wherein the event of the object refers to the object engaged
`in an activity.
`29. A non-transitory computer-readable storage medium
`containing instructions that when executed by a computer
`system cause said computer system to implementthe follow-
`ing method comprising:
`detecting first and second objects in a video from a single
`camera;
`detecting a plurality of attributes of each of the detected
`first and second objects by analyzing the video from said
`single camera, eachattribute representing a characteris-
`tic of the respective detected object;
`selecting a new user rule; and
`after detecting the plurality of attributes, identifying an
`eventthatis not one of the detected attributesofthefirst
`and second objects by applying the new userrule to the
`plurality of detected attributes, wherein the applying the
`new user rule to the plurality ofdetected attributes com-
`prises applving the new userrule to only the plurality of
`detected attributes;
`wherein the plurality of attributes that are detected are
`independent of which eventis identified,
`
`10
`
`15
`
`20
`
`25
`
`4
`wherein the step of identifying an event comprises identi-
`fying a first event ofthefirst object interacting with the
`second object by analyzing the detected attributes ofthe
`first and second objects, the first event not being one of
`the detected attributes, and
`wherein the event ofthe object refers to the object engaged
`in an activity.
`30. A video device comprising:
`meansfor detecting first and second objects in a video from
`a single camera;
`meansfor detecting a plurality of attributes ofthe object by
`analyzing the video from said single camera, each
`attribute representing a characteristic of the respective
`detected object;
`a memory storing the plurality of detected attributes; and
`meansfor identifying an event ofthefirst object interacting
`with the second object by applying a selected new user
`rule to the plurality of attributes stored in memory, and
`for identifying the event
`independent of when the
`attributes are stored in memory,the event not being one
`of the detected attributes,
`wherein the applying the selected new user rule to the
`plurality of attributes stored in memory comprises
`applying the selected new userrule to only the plurality
`ofattributes stored in memory,
`wherein the event ofthe object refers to the object engaged
`in an activity.
`
`Canon Ex. 1020 Page 3 of 3
`
`Canon Ex. 1020 Page 3 of 3
`
`