Set(findall(tp1.Parent, "Type", "line", "Tag", "tpTrackHistory_Tracks"). "Marker", ".", "MarkerSize",3, "MarkerEdgeColor",TrackColor, "MarkerFaceColor",TrackColor) TrackPlotter(tp1, "DisplayName", "Tracks", "HistoryDepth",TrackLength, "ConnectHistory", "off". Set(findall(a2, "Type", "line", "Tag", "tpTrackHistory_Truth"), "Color", "k") TrackPlotter(tp1, "DisplayName", "Truth", "HistoryDepth",0, "Marker", "^", "MarkerSize",4, "ConnectHistory", "off") H2 = uipanel(f, "FontSize",12, "Position", "Title", "Flock Center") Ī1 = axes(h1, 'Position',) H1 = uipanel(f, "FontSize",12, "Position", "Title", "Flock View") You can see how to generate code for a tracker in How to Generate C Code for a Tracker, and the tracker for this example is saved in the function flockTracker_kernel.mįunction = helperCreateDisplay(x,bound) The following line shows how the tracker is configured with the above properties. You want tracks to be quickly confirmed and deleted, and set the confirmation and deletion thresholds to and, respectively.įinally, you know that the sensor scans only a fraction of the flock at any given scan, and so you set the HasDetectableTrackIDsInput to true to be able to pass the detectable track IDs to the tracker. To limit the time required to calculate cost, you reduce the coarse cost calculation threshold in the AssignmentThreshold to a low value.įurther, you choose the more efficient Jonker-Volgenant as the assignment algorithm, instead of the default Munkres algorithm. You observe that the boids follow a curved path and choose a constant turn model defined by initctekf. You define the tracker as shown in the example How to Efficiently Track Large Numbers of Objects.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |