ippen / point-tracker

Interactive point tracking in videos, with options to export tracked points as JSON and create output videos.

Point Tracker

This repository contains a Python script for point tracking in videos using the Lucas-Kanade method. Users can select points interactively on the first frame of a video, and the script tracks these points across subsequent frames, visualizing them on the video. It also provides options to save the tracked points data to a JSON file and create an output video with tracked points.

Demo

langeoog_bimmelbahn_demo.mp4

Features

• Interactive point selection: Users can select points on the first frame of the video by clicking with the mouse.
• Visualization: Tracked points are visualized on each frame of the video.
• Output options: The script can create an output video with tracked points and save the tracked points data to a JSON file.
• Command-line interface: Easy-to-use command-line interface with options to specify input and output files.

Requirements

• Python 3.x
• OpenCV (cv2)
• NumPy

Usage

1. Clone the repository

   git clone https://github.com/ippen/point-tracker.git
   

2. Install the required dependencies using pip:

   pip install opencv-python numpy
   

3. Run the script with the following command:

   python point_tracker.py <video_path> [options]
   

   Replace <video_path> with the path to your input video file.

Command-line Options

• video_path: Path to the input video file.
• -s, --save: Optional. If set, saves tracked points to a JSON file (<video_name>_points.json) and the output video with tracked points to an MP4 file (<video_name>_tracked.mp4).

Example

python point_tracker.py input_video.mp4 --save

This command will track points in input_video.mp4 and save the tracked points data and the output video with tracked points if the --save flag is set.