Demo video: https://www.youtube.com/watch?v=-E1bOhiTmi0
The kalman filter is in motionModels.py. The tracker is in tracker.py,
Points are generated using constant velocity constant turning model.
This enviroment can generate non-stable update frequency, look at self.clock.tick() This enviroment can generate dropping observation and adding random noise to the observation. This enviroment can generate random false positive observation.
It also outputs the visualization of the enviroment using a pygame window
Red dots are the observation, white dots are the groud truth, other dots are the prediction
- scipy
- pygame, installation see:https://www.pygame.org/wiki/GettingStarted
open tracker.py, click run
Author: Zhihao