A Kinect-based Doppler-shift demonstration created by Seth Stubbs and Morgan Rehnberg.
Place a Kinect sensor around waist-height, with a display above. As users walk towards the display, the image will appear to be blue-shifted. As the user backs away, the image will appear red-shifted.
freenect(OpenKinect)python_libfreenect(freenect Python wrapper)OpenCv2&python_opencv2- Python 3.7
python3 pyshift.py <inputimage>
You can specify which image to display with the command line. If no input image is given, input.jpg in the current directory will be used.
There are a number of options that can be tweaked in pyshift.py
X_SCALE and Y_SCALE can both be used to stretch or downscale the image.
For example, if you have an image that is 1400x1200 you can scale it down by setting
X_SCALE = 0.5
Y_SCALE = 0.5so that the image is being rendered at 700x600, which should increase performance.
On less powerful systems (such as the Raspberry Pi) enabling the flash fix will result in a less-stuttery output.
FLASH_FIX = TrueThis should make a better effect when framerates are low.
This prints out some debug-related log information.
DEBUG = True