Abstract

LiDAR-assisted marker tracking helps the operator of a camera-based follow system keep the marker on a person more accurately, while the operator stays in control. In such systems an operator moves a marker over the image of a fixed camera, and the marker is converted into a stage position; the operator often misjudges depth in the flat video image and reacts with a delay. One or more LiDAR sensors calibrated to the same stage coordinates measure where persons are. The marker is pulled towards the nearby point cloud of a person like a magnet: mainly along the camera ray, which corrects the depth, and ahead in time by a prediction of the person's movement, which compensates for reaction and video delay. The pull acts only within a capture radius and gives way to the operator's input. The height of the output is the height of the stage surface under the marker, taken from a height model that can include moving platforms, plus an offset set by the operator. On a dark stage, a heatmap drawn over the camera image shows where movements take place that the camera cannot see; the heatmap fades over time, so that persons who have stopped stay visible as a fading trace. Heatmap and height work with or without the pull: the operator can switch the heatmap on to point in the dark by hand, and with the pull switched on, a marker placed on the heatmap is bound to the person.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS