[미국특허]
Method and system for motion vector-based video monitoring and event categorization
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/00
G06F-003/0481
H04L-029/06
H04N-021/431
H04N-021/433
H04N-021/4335
H04N-007/18
G06F-003/048
H04W-012/02
H04W-012/06
G06T-007/20
G06F-003/0484
G11B-027/00
G11B-027/028
H04L-029/08
G06F-003/0482
G11B-027/031
G11B-027/30
G11B-027/34
G06K-009/32
H04N-005/14
H04N-021/239
G11B-027/10
H04N-005/93
H04N-009/87
H04W-004/00
H04W-012/08
G08B-013/196
H04W-012/04
G06F-003/0485
G06F-003/0488
H04N-021/2187
H04N-021/2743
H04N-021/462
H04N-021/422
출원번호
US-0821597
(2015-08-07)
등록번호
US-9886161
(2018-02-06)
발명자
/ 주소
Laska, Jason N.
Nelson, Gregory R.
Duffy, Greg
출원인 / 주소
GOOGLE LLC
대리인 / 주소
Morgan, Lewis & Bockius LLP
인용정보
피인용 횟수 :
0인용 특허 :
85
초록▼
A computer system processes a video stream to detect a start of a first motion event candidate in the video stream, and in response to detecting the start of the first motion event candidate in the video stream, initiates event recognition processing on a first video segment associated with the star
A computer system processes a video stream to detect a start of a first motion event candidate in the video stream, and in response to detecting the start of the first motion event candidate in the video stream, initiates event recognition processing on a first video segment associated with the start of the first motion event candidate. Initiating the event recognition processing further includes: determining a motion track of a first object identified in the first video segment; generating a representative motion vector for the first motion event candidate based on the motion track of the first object; and sending the representative motion vector for the first motion event candidate to an event categorizer, where the event categorizer assigns a respective motion event category to the first motion event candidate based on the representative motion vector of the first motion event candidate.
대표청구항▼
1. A method of processing a video stream, comprising: processing the video stream to detect a start of a first motion event candidate in the video stream, wherein processing comprises: obtaining a profile of a motion pixel count for a current frame sequence in the video stream;in response to determi
1. A method of processing a video stream, comprising: processing the video stream to detect a start of a first motion event candidate in the video stream, wherein processing comprises: obtaining a profile of a motion pixel count for a current frame sequence in the video stream;in response to determining that the obtained profile meets a predetermined trigger criterion, determining that the current frame sequence includes a motion event candidate;identifying a beginning time for a portion of the profile meeting the predetermined trigger criterion; anddesignating the identified beginning time to be the start of the first motion event candidate; andin response to detecting the start of the first motion event candidate in the video stream, initiating event recognition processing on a first video segment associated with the start of the first motion event candidate. 2. The method of claim 1, wherein determining that the obtained profile meets the predetermined trigger criterion includes determining that the motion pixel count satisfies a threshold motion pixel count. 3. The method of claim 1, wherein the start of the first motion event candidate is the time at which the motion pixel count meets the predetermined trigger criterion. 4. The method of claim 1, wherein determining that the obtained profile meets the predetermined trigger criterion is with respect to a predetermined length of the current frame sequence. 5. The method of claim 1, further comprising adjusting the predetermined trigger criterion over time based on a performance feedback. 6. The method of claim 5, wherein the performance feedback indicates a number of false-positive motion events detected in processing the video stream. 7. The method of claim 6, wherein the adjusting comprises increasing the predetermined trigger criterion when the number of false positives satisfies a corresponding threshold. 8. The method of claim 5, wherein the performance feedback indicates a number of motion events that failed to be detected in processing the video stream. 9. The method of claim 8, wherein the adjusting comprises decreasing the predetermined trigger criterion when the number of motion events that failed to be detected satisfies a corresponding threshold. 10. A computing system for processing a video stream, comprising: one or more processors; andmemory having instructions stored thereon, the instructions, when executed by the one or more processors, cause the processors to perform operations comprising:processing the video stream to detect a start of a first motion event candidate in the video stream, wherein processing comprises: obtaining a profile of a motion pixel count for a current frame sequence in the video stream;in response to determining that the obtained profile meets a predetermined trigger criterion, determining that the current frame sequence includes a motion event candidate;identifying a beginning time for a portion of the profile meeting the predetermined trigger criterion; anddesignating the identified beginning time to be the start of the first motion event candidate; andin response to detecting the start of the first motion event candidate in the video stream, initiating event recognition processing on a first video segment associated with the start of the first motion event candidate. 11. The computing system of claim 10, wherein the operations further comprise adjusting the motion pixel count based on a camera state change that causes a change in pixel values in the current frame sequence, wherein the change in pixel values does not correspond to the occurrence of motion in the current frame sequence. 12. The computing system of claim 11, wherein the operations further comprise suppressing designation of the identified beginning time as the start of the first motion event candidate in accordance with the start of the first motion event candidate overlapping with a time corresponding to the camera state change. 13. The computing system of claim 11, wherein adjusting the motion pixel count comprises adjusting the motion pixel count by a fraction. 14. The computing system of claim 11, wherein adjusting the motion pixel count comprises resetting the motion pixel count after the camera state change. 15. The computing system of claim 10, wherein processing the video stream further comprises: detecting a rate of change for the motion pixel count of the profile; andif the detected rate of change satisfies a predetermined rate, suppressing designation of the identified beginning time as the start of the first motion event candidate. 16. The computing system of claim 15, wherein the operations further comprise, after the suppressing: executing a recovery process to determine whether the identified beginning time corresponds to a valid motion event candidate; andin accordance with determining that the identified beginning time corresponds to a valid motion event candidate, recovering the first motion event candidate by designating the identified beginning time to be the start of the first motion event candidate. 17. The computing system of claim 15, wherein the operations further comprise resetting the motion pixel count if the detected rate of change satisfies the predetermined rate. 18. The method of claim 1, wherein initiating event recognition processing on the first video segment associated with the start of the first motion event candidate further comprises: determining a motion track of a first object identified in the first video segment; andgenerating a representative motion vector for the first motion event candidate based on the motion track of the first object, wherein a respective motion event category is assigned to the first motion event candidate based on the representative motion vector of the first motion event candidate;wherein determining the motion track of the object identified in the first video segment further comprises, based on a frame sequence of the first video segment: building a histogram that specifies a frame count for pixel locations in a scene of the first video segment;segmenting the histogram into one or more motion regions; andselecting one or more dominant motion regions from the one or more motion regions based on a predetermined dominance criterion, wherein at least one of the one or more dominant motion regions corresponds to the respective motion track of the object. 19. The method of claim 18, wherein the frame count for a respective pixel location is a sum of the motion pixel count at the respective pixel location in frames of the first video segment. 20. The method of claim 19, further comprising creating an event mask that includes one or more of the pixel locations having less than a threshold motion pixel count. 21. The method of claim 20, wherein the threshold motion pixel count is with respect to a predefined number of frames of the first video segment. 22. The method of claim 20, wherein one or more of the pixel locations not included in the event mask include at least the one or more dominant motion regions. 23. The method of claim 20, wherein the event mask is a non-binary mask. 24. The method of claim 18, wherein the predetermined dominance criterion is a threshold frame count. 25. The method of claim 18, wherein the predetermined dominance criterion is a threshold motion pixel count. 26. The computing system of claim 10, wherein initiating event recognition processing on the first video segment associated with the start of the first motion event candidate further comprises: determining a motion track of a first object identified in the first video segment; andgenerating a representative motion vector for the first motion event candidate based on the motion track of the first object, wherein a respective motion event category is assigned to the first motion event candidate based on the representative motion vector of the first motion event candidate;wherein determining the motion track of the object identified in the first video segment further comprises, based on a frame sequence of the first video segment: building a histogram that specifies a frame count for pixel locations in a scene of the first video segment;segmenting the histogram into one or more motion regions; andselecting one or more dominant motion regions from the one or more motion regions based on a predetermined dominance criterion, wherein at least one of the one or more dominant motion regions corresponds to the respective motion track of the object. 27. The computing system of claim 26, wherein the operations further comprise determining a track length of the motion track. 28. The computing system of claim 27, wherein the operations further comprise suppressing the motion track of the first object in accordance with the determined track length satisfying a predetermined threshold. 29. The computing system of claim 28, wherein the suppression of the motion track occurs before generating the representative motion vector, and wherein the operations further comprise foregoing generation of the representative motion vector in accordance with suppression of the motion track. 30. The computing system of claim 26, wherein determining the motion track of the object further comprises performing object segmentation to identify one or more foreground objects in the first video segment. 31. The computing system of claim 30, wherein the one or more foreground objects identified in the first video segment include the object for which the motion track is determined.
Saft, Keith D.; Bernoulli, Carlo P.; Parry, Lee J.; Shephard, Megan D.; Poelker, Cole J., Graphic user interface for displaying content selections on a display panel.
Saft, Keith D.; Bernoulli, Carlo P.; Parry, Lee J.; Shephard, Megan D.; Poelker, Cole J., Graphical user interface for displaying content selections on a display panel.
Desimone, Michael J.; Hampapur, Arun; Lu, Zuoxuan; Mercier, Carl P.; Milite, Christopher S.; Russo, Stephen R.; Shu, Chiao-Fe; Tan, Chek K., Identifying spatial locations of events within video image data.
Kawashima, Yuji; Kikuchi, Yoshihiro; Fujisawa, Tatsuro; Suzuki, Shingo, Information processing apparatus with video encoding process control based on detected load.
Laska, Jason N.; Nelson, Gregory R.; Duffy, Greg; Mitsuji, Hiro; Hill, Cameron; Davidsson, Martin; Montalbo, Michael D.; Wan, Tung Yuen, Method and system for cluster-based video monitoring and event categorization.
Laska, Jason N.; Nelson, Greg R.; Duffy, Greg; Hill, Cameron; Davidsson, Martin, Method and system for retroactively changing a display characteristic of event indicators on an event timeline.
Sharma, Rajeev; Mummareddy, Satish; Hershey, Jeff; Jung, Namsoon, Method and system for segmenting people in a physical space based on automatic behavior analysis.
Schonfeld,Dan; Hariharakrishnan,Karthik; Raffy,Philippe; Yassa,Fathy, Occlusion/disocclusion detection using K-means clustering near object boundary with comparison of average motion of clusters to object and background motions.
Borzycki, Andrew; Deva, Mallikharjuna Reddy; Gajendar, Uday Nandigam; Roychoudhry, Anil, Single sign-on access in an orchestration framework for connected devices.
Lane, Corey A.; Buck, Heidi L.; Li, Joshua S.; Bagnall, Bryan D.; Stastny, John C.; Hallenborg, Eric C., System for tracking maritime domain targets from full motion video.
Pantofaru, Caroline Rebecca; Bettadapura, Vinay; Bharat, Krishna; Essa, Irfan, Systems and methods for attention localization using a first-person point-of-view device.
Laska, Jason N.; Hua, Wei; Chaudhry, Rizwan Ahmed; Varadharajan, Srivatsan; Heitz, III, George Alban, Systems and methods for categorizing motion event candidates.
Watts, Tim J.; Offerdahl, Alex; Atwood, Joe; Driscoll, Jim; Loar, Steve; Stabnow, Jeff, Systems, computer-implemented methods, and computer medium to determine premiums and indemnities for supplemental crop insurance.
Tangeland, Kristian; Aarrestad, Glenn R. G.; Nielsen, Johan Ludvig; Stuan, Øivind, Using local talker position to pan sound relative to video frames at a remote location.
Wilson Charles Park ; Pedersen ; Jr. Chris Harvey ; Auyeung Alex Kamlun ; MacCormack David Ross, Video data capture and formatting in intelligent video information management system.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.