shomnathsomu/crack-detection-opencv ? reverse-engineered prompt

Build me a Python app that detects cracks in road or concrete surface photos using OpenCV.

I want to be able to drop a few input images into an input folder, run one script, and get processed images plus final crack detection results saved into output folders. The pipeline should feel like the one in the project, starting with grayscale and blur, then a log style enhancement, then smoothing, edge detection, and a closing step to connect broken crack lines. After that, highlight the detected crack areas clearly on the original image with feature points or markings so it is easy to see what the app found.

Please make it work with the sample images in the repo, keep it simple to run from the command line, and organize the code so it is easy to tweak the image processing settings later. If anything is unclear, look up current OpenCV docs online and make sensible choices.