# 05_disparity/libsgm — dependencies
#
# The Python pipeline runner (run_sgm_pipeline.py) uses only the standard library.
# Disparity computation is done by the compiled stereosgm_new binary (CUDA C++).
#
# --- System build requirements (not installable via pip) ---
#   - NVIDIA GPU with CUDA compute capability >= 3.5
#   - CUDA Toolkit 11.x or 12.x
#   - CMake >= 3.18
#   - OpenCV (C++ headers + libs, for building libSGM samples)
#
# Build:
#   cd ~/Speckle-Scanner/05_disparity/libsgm
#   mkdir -p build && cd build
#   cmake .. -DENABLE_SAMPLES=on
#   make stereosgm_new -j4
#
# No pip packages required to run run_sgm_pipeline.py after the binary is built.