Multiple Object Tracking Algorithm Using Kalman Filter.

A Python package for multi-object detection.

Press to watch the video:

Dependencies

Circles tracking

• python 3.x
• NumPy
• sklearn
• opencv

Object tracking

• python3.x
• NumPy
• sklearn
• OpenCV
• tensorflow2

Execution

Circles tracking

Press to watch the video:

Create and track data simultaneously

This option creates random sample data while tracking the data simultaneously.

To run:

python createAndTrackSimpleData.py --ball {Number of circles to track}

• ball: int. Number of balls to track. The default value is set to 5.

Create data

This option creates a short video of multiple circles randomly moving.

To run:

python createSimpleData.py --name {Name of the file} --ball {Number of circles to track}

• name: str. The name of the video this code creates. Defalut name is "ex5". The file will be saved in the video folder in "avi" format.
• ball: int. Number of balls to track. The default value is set to 5.

Track data

This option opens a short video of multiple circles randomly moving and tracking them.

To run:

python simpleMultiTracker.py --name {Name of the file} --save {True or False} 

• name: str. The name of the video this code reads. Defalut name is "ex5.avi".
• save: bool. True for saving the file and False for not. The default value is set to False. if set to True, saves the file in the "savedVideo" folder.

Object tracking

Pre-Execution step

• This package uses a yolov3 model. The yolov3 model was generated following this tutorial:

implementation-of-yolov3-simplified

• The model's name is "yolo_darknet" and is saved in the "model" folder.
• The classes' names are located in the "data" folder.

In the case of using another model, change MODEL_SIZE (image input size) values in "yoloDetector.py" to the relevant ones.

Track data

This option opens a short video of multiple moving objects and tracks them.

To run:

python yoloMultiTracker.py --VideoName {Name of the file} --save {True or False} 

• name: str. The name of the video this code reads. Defalut name is "video/1.mp4".
• save: bool. "True" for saving the file and "False" for not. The default value is set to False. if set to True, saves the file in the "savedVideo" folder.