Rodent eyeblink detector for classical eyeblink conditioning. Used for high speed videos captured from Basler's acA640-750um. Contains custom video recorder for Basler USB-cameras (may work with others too) made with PyPylon. Image processing done with OpenCV and GUI with python's Tkinter.

Note: Program is at initial prototype state, but still usable.

• Detect eye with haar cascade. Cascade from human eye, but works well with animals too.
• Tracking eye of the moving animal (experimental)
• Frame Filtering (brightness, contrast, blur)
• Frame thresholding (binary image)
• Selects ROI area
• CSV-output for frame, time and white level from binarized frames
• Load and save parameters
• Command line interface
• Quick and dirty GUI

• Set output file (avi)
• Select video compression via Video for Windows interface. Not tested with UNIX-like OS:s.
• Set Basler camera parameters (width, height, fps, exposure time)
• Preview camera output (still frame).
• Record

• Docstrings
• Save overwrite check/message
• Threading/multiprocessing for better performance
• Real-time plotting
• Peak detection (realtime)
• Rework to meet the MVC pattern requirements for better structure and maintainability
• Better GUI (code and ui)
• Refactoring and cleaning

• Docstrings
• Threading/multiprocessing for better performance
• Rework to meet the MVC pattern requirements for better structure and maintainability
• Better GUI (code and ui)
• Video preview (instead of still frame)
• Refactoring and cleaning

• Install Python 3 (Miniconda)
• Create Conda environment. if you wish.
• Activate environment..
• PIP install following packages:
  • numpy, opencv-python, opencv-contrib-python, Pillow, scikit-video
• In addition for video recorder install Basler Pylon SDK and pip pypylon

To run gui:

python blinky.py

To run cmd line see the help:

python blinky.py -help

To run gui:

python record.py