QGIS plugin for analysing ground-based oblique images
Reference: Wright, C., Bone, C., Mathews, D., Tricker, J., Wright, B., and Higgs, E. (2024) Mountain Image Analysis Suite (MIAS): A new plugin for converting oblique images to landcover maps in QGIS. Transactions in GIS.doi.org/10.1111/tgis.13229
This plugin contains four tools for analysing ground-based oblique images. The final product is a classified and spatially referenced viewshed representing the landscape shown in the photograph. MIAS harnesses PyLC (Python Landscape Classifier) available independently here. The training dataset is sampled from the Mountain Legacy Project repeat photography collection hosted at the University of Victoria.
- numpy >=1.18.5
- opencv >=3.4.1
- torch >=1.6.0
- scikit-image >=0.19.3
Install the latest version of QGIS through the OSGeo4W installer from here. From the OSGeo4W installer, select Express install. Choose QGIS LTR, GDAL, and GRASS GIS from the option menu.
Install the latest version of QGIS here.
MIAS relies on some Python packages that do not come installed with QGIS and has conflicts with the existing versions of opencv and numpy. From the Plugins menu on QGIS, open the Python console and type the following commands:
import pip
pip.main(['uninstall','-y','opencv-contrib-python'])
pip.main(['install','opencv-python'])
pip.main(['install','--upgrage','numpy'])
pip.main(['install','torch'])
pip.main(['install','scikit-image'])From the Code menu (green button) on the GitHub page, select Download ZIP.
From the Plugins menu in QGIS, choose Manage and Install Plugins, then Install from ZIP. Upload the ZIP file that you just downloaded from GitHub.
A video tutorial for MIAS is available here. Example data and written instructions can be downloaded here.