The chapter introduces inverse rendering and the role of image features in 3D reconstruction. The example here walks through Harris corner detection, from image gradients to the final corner locations. It runs on the CPU.
With Python installed, run these commands from Chapter01:
python -m pip install -r requirements.txt
python HarrisCornerDetection.pyAt the prompt, press Enter to use the generated checkerboard, enter
testImg.jpg for the bundled photo, or give the path to your own image.
A Matplotlib window shows the input, horizontal and vertical gradients,
gradient strength, Harris response, and detected corners. To experiment,
change k, threshold, or neighborhood_size in the call to
harris_corner_detector() near the end of the script.