Skip to content

Latest commit

 

History

History
43 lines (27 loc) · 1.66 KB

README.md

File metadata and controls

43 lines (27 loc) · 1.66 KB
Raw

Trajectory Estimation

This add-on calculates the direction of movement of trackable objects based on their past and present coordinates in the frame.

Config

The class TrajectoryProcessor is initialized with the following config parameters:

DEFAULT = {
    "centroid_index": 2,
    "temporal_length": 10
}

where:

  • centroid_index (int): -nth centroid to compare with the current centroid. This comparison occurs when we don't have enough centroid datapoint as the temporal_length requires
  • temporal_length (int): The amount of centroid points to compare against the present centroid position.

Input

The post_process gets as input the AddonObject and relies on the results of the tracking add-on set in:

  • AddonObject.shared['trackable_objects'] (format: OrderedDict)
  • AddonObject.shared['trackable_objects_history'] (format: OrderedDict)

Direct dependency with the tracking add-on.

Process

After object detection and object tracking, the trajectory estimator receives the full history of trackable_objects from the tracking add-on. For every trackable_object we check its previous centroid positions from the past frames and estimate the object's direction of movement on the current frame.

We support the following directions format:

'up', 'down', 'left', 'right', 'downleft', 'downright', 'upleft', 'upright' or '' if direction can't be set.

Output

The add-on's output is shared on the AddonObject.inference.extra for each object that was actively tracked by the tracking add-on on the current frame:

  • AddonObject.inference.extra['movement_directions'] - (format: dict where {'obj_id': direction})