README.md

metadata

pretty_name: Monte Maps
license: mit
task_categories:
  - other
tags:
  - 3d

  __  __    U  ___ u  _   _     _____  U _____ u      __  __      _       ____     ____     
U|' \/ '|u   \/"_ \/ | \ |"|   |_ " _| \| ___"|/    U|' \/ '|uU  /"\  u U|  _"\ u / __"| u  
\| |\/| |/   | | | |<|  \| |>    | |    |  _|"      \| |\/| |/ \/ _ \/  \| |_) |/<\___ \/   
 | |  | |.-,_| |_| |U| |\  |u   /| |\   | |___       | |  | |  / ___ \   |  __/   u___) |   
 |_|  |_| \_)-\___/  |_| \_|   u |_|U   |_____|      |_|  |_| /_/   \_\  |_|      |____/>>  
<<,-,,-.       \\    ||   \\,-._// \\_  <<   >>     <<,-,,-.   \\    >>  ||>>_     )(  (__) 
 (./  \.)     (__)   (_")  (_/(__) (__)(__) (__)     (./  \.) (__)  (__)(__)__)   (__)      

MonteMaps is a Google Maps-like software that allows the user to localize themselves in the Montefiore Institute (ULiege) using a off-the-self camera. The code can be obtained on GitHub.

Content

• model/: the building model and camera poses in COLMAP format.
• reconstructio-video-images-10fps.zip: the frames extracted at 10 fps from the building's recording used for the reconstruction, the camera poses estimation and VPR database.
• megaloc_descriptors.npz: the global descriptors of the extracted frames, generated with MegaLoc.
• test-sequence-obstruction.zip: a test video sequence of 546 images in Montefiore with obstruction.
• test-sequence-no-obstruction.zip: a test video sequence of 680 images in Montefiore without obstruction.
• test-images.zip: a serie of 46 test images with/without obstruction, in day/night condition, with/without blur
• buste.glb: a 3d model used as position marker in the renderer of the project (created using SAM3D).
• points_of_intereset.json: list of points of interesets as key(camera ID) - values(name)

Note: reconstructio-video-images-10fps is recorded on DJI Osmo-3, test-images.zip is recorded on iPhone 13 Mini