Location-based signature selection for multi-camera object tracking
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/00
G06K-009/62
G06T-007/20
G06K-009/32
출원번호
US-0078292
(2013-11-12)
등록번호
US-9524448
(2016-12-20)
우선권정보
AU-2009-243528 (2009-12-04)
발명자
/ 주소
Wedge, Daniel John
출원인 / 주소
CANON KABUSHIKI KAISHA
대리인 / 주소
Fitzpatrick, Cella, Harper & Scinto
인용정보
피인용 횟수 :
0인용 특허 :
15
초록▼
Disclosed herein are a method, system, and computer program product for determining a correspondence between a first object (713) tracked in a first field of view and a second object tracked (753) in a second field of view. The method determines a first area (711) in the first field of view, based o
Disclosed herein are a method, system, and computer program product for determining a correspondence between a first object (713) tracked in a first field of view and a second object tracked (753) in a second field of view. The method determines a first area (711) in the first field of view, based on the location and size of the first object (713). The method utilizes a predetermined area relationship between the first area (711) in the first field of view and at least one area (751) in the second field of view to determine a second area (751) in the second field of view. In one embodiment, the method determines the second area (751) in the second field of view by comparing predetermined area relationships between the first area (711) and any areas (751) in the second field to determine a best match. The method determines a correspondence between the first object (713) and the second object (753), based on a comparison between a first object signature associated with the first object (713) and a second object signature associated with the second object (753).
대표청구항▼
1. A method of linking tracks of objects in different fields of view, comprising: detecting a first object in a first field of view and a second object in a second field of view;generating a first track and a second track by tracking the first object in the first field of view and the second object
1. A method of linking tracks of objects in different fields of view, comprising: detecting a first object in a first field of view and a second object in a second field of view;generating a first track and a second track by tracking the first object in the first field of view and the second object in the second field of view, respectively;accumulating, during the tracking, a plurality of first signatures representing respective appearances of the first object corresponding to different views of the first object in association with the first track and a plurality of second signatures representing respective appearances of the second object corresponding to different views of the second object in association with the second track;determining a first area in the first field of view, based on a portion of the first object in a frame associated with the first track;determining a plurality of second areas corresponding to the first area in the second field of view in a plurality of frames associated with the second track, by comparing a first one of the first signatures of the first track corresponding to the first area with the plurality of second signatures of the second track;comparing the first one of the first signatures corresponding to the first area and a respective one of the second signatures corresponding to the second areas; andlinking the first track and the second track, if the first one of the first signatures matches with any of the second signatures corresponding to the second areas as a result of the comparison. 2. The method according to claim 1, wherein each of the first and the second signatures includes a predetermined number of signature modes. 3. The method according to claim 1, further comprising: passing a signature between the first track associated with the first field of view and the second track associated with the second field of view. 4. The method according to claim 3, where an order of passing the signature is prioritized based on at least one of a position of said first area in the first field of view, a position of said second area in the second field of view, and a number of previous matches relating to each signature mode. 5. The method according to claim 1, wherein said comparing includes: determining a time of transit between the second area in the second field of view and the first area in the first field of view, andusing the time of transit in said comparison. 6. The method according to claim 1, wherein each of said first signature and said second signature has at least one signature mode, wherein each signature mode represents one view of an object. 7. The method according to claim 6, wherein each of said first signature and said second signature is associated with a predetermined maximum number of signature modes. 8. The method according to claim 1, wherein said determining of said second areas is based on at least one characteristic of a second object. 9. The method according to claim 8, wherein each of said at least one characteristic is one of a spatial, temporal, and appearance characteristic associated with said first object. 10. The method according to claim 1, further comprising: determining whether there are any area relationships between the first area and the area in the second field of view; andcreating a new area relationship between the first area and the area in the second field of view if there are no area relationships therebetween. 11. The method according to claim 1, wherein the accumulating the plurality of first signatures includes: comparing a current appearance of the first object with each of existing first signatures; andadding the current appearance to the first signatures as a new signature, if a difference between the current appearance and each of the existing first signatures is larger than a threshold. 12. The method according to claim 11, wherein the accumulating the plurality of first signatures further includes: updating a specific one of the first signatures based on the current appearance, if a difference between the current appearance and the specific one of the first signatures is not larger than the threshold. 13. The method according to claim 1, further comprising: transferring one of the first signatures, which is not included in the second signatures, to the second signatures if the first track and the second track have been linked. 14. An apparatus for linking tracks of objects in different fields of view, said apparatus comprising: a storage device for storing a computer program; anda processor for executing the program, said program comprising code for performing the steps of:detecting a first object in a first field of view and a second object in a second field of view;generating a first track and a second track by tracking the first object in the first field of view and the second object in the second field of view, respectively;accumulating, during the tracking, a plurality of first signatures representing respective appearances of the first object corresponding to different views of the first object in association with the first track and a plurality of second signatures representing respective appearances of the second object corresponding to different views of the second object in association with the second track;determining a first area in the first field of view, based on a portion of the first object in a frame associated with the first track;determining a second plurality of second areas corresponding to the first area in the second field of view in a plurality of frames associated with the second track, by comparing a first one of the first signatures of the first track corresponding to the first area with the plurality of second signatures of the second track; andcomparing the first one of the first signatures to the first area and a respective one of the second signatures corresponding to the second areas;linking the first track and the second track, if the first one of the first signatures matches with any one of the second signatures corresponding to the second areas as a result of the comparison. 15. A non-transitory computer readable storage medium having recorded thereon a computer program for directing a computer to execute a method of linking tracks of objects in different fields of view, said computer program comprising code for performing the steps of: detecting a first object in a first field of view and a second object in a second field of view;generating a first track and a second track by tracking the first object in the first field of view and the second object in the second field of view, respectively;accumulating, during the tracking, a plurality of first signatures representing respective appearances of the first object corresponding to different views of the first object in association with the first track and a plurality of second signatures representing respective appearances of the second object corresponding to different views of the second object in association with the second track;determining a first area in the first field of view, based on a portion of the first object in a frame associated with the first track;determining a plurality of second areas corresponding to the first area in the second field of view in a plurality of frames associated with the second track, by comparing a first one of the first signatures of the first track corresponding to the first area with the plurality of second signatures of the second track;comparing the first one of the first signatures to the first area and a respective one of the second signatures corresponding to the second areas; andlinking the first track and the second track, where the first one of the first signatures matches with any one of the second signatures corresponding to the second areas as a result of the comparing.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (15)
Haupt, Gordon T.; Fleischer, Stephen D.; Vallone, Robert P.; Russell, Stephen G.; Frederick, Timothy B., Automated searching for probable matches in a video surveillance system.
Haupt, Gordon T.; Freeman, J. Andrew; Fleischer, Stephen D.; Vallone, Robert P.; Russell, Stephen G.; Frederick, Timothy B., Interactive system for recognition analysis of multiple streams of video.
Zhao, Tao; Aggarwal, Manoj; Kumar, Rakesh; Sawhney, Harpreet, Method and apparatus for tracking objects over a wide area using a network of stereo sensors.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.