Stanford EE 392J - Multiple Camera Object Tracking

Unformatted text preview:

Multiple Camera Object TrackingOutlineObject TrackingSingle Object & Single CameraSingle Object & Multiple CameraSystem ArchitectureStatic Point CorrespondenceDynamic Point CorrespondenceBlock-Based Motion EstimationAdaptive Window SizingFour Corner MethodExample: Four Corner MethodCorrelative MethodCorrelative Method (con’t)Example: Correlative MethodOcclusion DetectionRedetection Procedure (1 Camera)Example: OcclusionCamera CommunicationResultConclusionEE392J Final Project, March 20, 20021Multiple Camera Object TrackingMultiple Camera Object TrackingHelmy Eltoukhy and Khaled SalamaHelmy Eltoukhy and Khaled SalamaEE392J Final Project, March 20, 20022OutlineOutlineIntroductionIntroductionPoint Correspondence between multiple Point Correspondence between multiple camerascamerasRobust Object TrackingRobust Object TrackingCamera Communication and decision Camera Communication and decision makingmakingResultsResultsEE392J Final Project, March 20, 20023Object TrackingObject TrackingThe objective is to obtain an accurate The objective is to obtain an accurate estimate of the position (x,y) of the object estimate of the position (x,y) of the object trackedtrackedTracking algorithms can be classified intoTracking algorithms can be classified into•Single object & Single Camera Single object & Single Camera •Multiple object & Single Camera Multiple object & Single Camera •Multiple objects & Multiple CamerasMultiple objects & Multiple Cameras•Single object & Multiple CamerasSingle object & Multiple CamerasEE392J Final Project, March 20, 20024Single Object & Single CameraSingle Object & Single CameraAccurate camera calibration and scene modelAccurate camera calibration and scene modelSuffers from OcclusionsSuffers from OcclusionsNot robust and object dependantNot robust and object dependantEE392J Final Project, March 20, 20025Single Object & Multiple CameraSingle Object & Multiple CameraAccurate point correspondence between scenesAccurate point correspondence between scenesOcclusions can be minimized or even avoidedOcclusions can be minimized or even avoidedRedundant information for better estimationRedundant information for better estimationMultiple camera Communication problemMultiple camera Communication problemEE392J Final Project, March 20, 20026System ArchitectureSystem ArchitectureObjectIdentificationObjectTrackingCheck Position(X1-X2) <(Y1-Y2) <ChooseCameraViewObjectIdentificationObjectTrackingEE392J Final Project, March 20, 20027Static Point CorrespondenceStatic Point CorrespondenceThe output of the tracking stage isThe output of the tracking stage isA simple scene model is used to get real A simple scene model is used to get real estimate of coordinates estimate of coordinates Both Affine and Perspective models were Both Affine and Perspective models were used for the scene modeling and static used for the scene modeling and static corresponding points were used for corresponding points were used for parameter estimationparameter estimationLeast mean squares was used to improve Least mean squares was used to improve parameter estimationparameter estimation)(),( nYnXii)(ˆ),(ˆnYnXiiEE392J Final Project, March 20, 20028Dynamic Point CorrespondenceDynamic Point CorrespondenceAffine modelusing A(n)B(n)Affine modelusing A(n)Add this point to AAdd this point to ACheck Position(X1-X2) < T(Y1-Y2) < TEE392J Final Project, March 20, 20029Block-Based Motion EstimationBlock-Based Motion EstimationTypically, in object tracking precise sub-pixel Typically, in object tracking precise sub-pixel optical flow estimation is not needed.optical flow estimation is not needed.Furthermore, motion can be on the order of Furthermore, motion can be on the order of several pixels, thereby precluding use of gradient several pixels, thereby precluding use of gradient methods.methods.We started with a simple sum of squared We started with a simple sum of squared differences error criterion coupled with full search differences error criterion coupled with full search in a limited region around the tracking window.in a limited region around the tracking window.• 2),()),,(),,(( tyxsttnymxsSSDcyxcerrorEE392J Final Project, March 20, 200210Adaptive Window SizingAdaptive Window SizingAlthough simple block-based motion Although simple block-based motion estimation may work reasonably well when estimation may work reasonably well when motion is purely translational, it can lose the motion is purely translational, it can lose the object if its relative size changes.object if its relative size changes.•If the object’s camera field of view shrinks, the If the object’s camera field of view shrinks, the SSD error is strongly influenced by the SSD error is strongly influenced by the background.background.•If the object’s camera field of view grows, the If the object’s camera field of view grows, the window fails to make use of entire object window fails to make use of entire object information and can slip away. information and can slip away.EE392J Final Project, March 20, 200211Four Corner MethodFour Corner MethodThis technique divides the rectangular object This technique divides the rectangular object window into 4 basic regions - each one quadrant.window into 4 basic regions - each one quadrant.Motion vectors are calculated for each subregion Motion vectors are calculated for each subregion and each controls one of four corners.and each controls one of four corners.Translational motion is captured by all four moving Translational motion is captured by all four moving equally, while window size is modulated when equally, while window size is modulated when motion is differential.motion is differential.Resultant tracking window can be non-rectangular, Resultant tracking window can be non-rectangular, i.e., any quadrilateral approximated by four i.e., any quadrilateral approximated by four rectangles with a shared center corner.rectangles with a shared center corner.EE392J Final Project, March 20, 200212Example: Four Corner MethodExample: Four Corner MethodSynthetically generated test sequences:EE392J Final Project, March 20, 200213Correlative MethodCorrelative MethodFour corner method is strongly subject to error Four corner method is strongly subject to error accumulation which can result in drift of one or accumulation which can result in drift of one or more of the tracking


View Full Document

Stanford EE 392J - Multiple Camera Object Tracking

Download Multiple Camera Object Tracking
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Multiple Camera Object Tracking and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Multiple Camera Object Tracking 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?