[PYTHON] Run OpenMVG on Mac
Run OpenMVG on Mac. This time, we will run the tutorial that is installed by default and display the results.
Execution environment
macOS Mojave (version 10.14.5) pyenv+python 3.7.5 cmake 3.16.4 Xcode 11.3.1
procedure
Install OpenMVG
Install the required python library
$ pip install sphinx
Clone source code from Github
$ git clone --recursive https://github.com/openMVG/openMVG.git
Edit openMVG / src / CMakeLists.txt Before editing:
# ==============================================================================
# IMAGE IO detection
# ==============================================================================
find_package(JPEG QUIET)
find_package(PNG QUIET)
find_package(TIFF QUIET)
After editing: (Add set ~)
# ==============================================================================
# IMAGE IO detection
# ==============================================================================
set(CMAKE_FIND_FRAMEWORK LAST)
find_package(JPEG QUIET)
find_package(PNG QUIET)
find_package(TIFF QUIET)
Build using Xcode
$ mkdir openMVG_Build
$ cd openMVG_Build
$ cmake -DCMAKE_BUILD_TYPE=RELEASE -G "Xcode" . ../openMVG/src/
$ xcodebuild -configuration Release
Run the tutorial
Run the tutorial
$ cd software/SfM
$ python3 tutorial_demo.py
In the tutorial, a 3D point cloud will be generated from the 10 images stored in the following folder.
openMVG_Build/software/SfM/ImageDataset_SceauxCastle/images
The internal parameters of the camera are also stored
$ cat K.txt
2905.88 0 1416
0 2905.88 1064
0 0 1
Install MeshLab
Download and install the mac version from the MeshLab site https://www.meshlab.net/
From [File]> [Import Mesh.] openMVG_Build/src/software/SfM/tutorial_out/reconstruction_sequential/colorized.ply Read
If it is displayed as shown below, it is successful.
Recommended Posts