v1.1.0
This release fixed issue #2. Adding support for additional landmarks information for easier mapping.
The update
method expect a single arguments that had a the format List[List[float]]
, which means a list of detected object in that frame. Each object will have the format: [x_top, y_top, width, height, confidence]
or [x_top, y_top, width, height, confidence, landmark_x1, landmark_y1, ...]
for additional landmark associated with each bounding box.
The update method will return a list of tracked object in the format List[Object]
, each object will have the following structure:
{
bbox: List[(int) x_top, y_top, width, height],
track_id: int,
landmarks: List[(float) x1, y1, x2, y2, ..., x_n, y_n],
}