Skip to content

PerceptionRobotique/VisualAlignment

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Visual Alignment

Direct visual pose estimation of omnidirectional images

Based on [LibPeR] C++ library, this repository contains code to demonstrate use cases of direct visual orientation estimation. Use MPP_SSD_alignment_VVSlike_EquiRect. MPP_SSD_alignment_EquiRect is let as legacy for the moment.

How to use

Build

  1. install libPeR, libboost-filesystem-dev, libboost-system-dev, libboost-regex-dev
  2. create a new directory named build in MPP_SSD_alignment_VVSlike_EquiRect
  3. use cmake to fill the build directory
  4. open the project in build or use make in the latter directory to build the exe file
  5. run the program from the command line at the project top directory, considering it includes the 2021_MPP_SSD_alignment_EquiRect_media directory, with arguments as:

./build/MPPSSDalignment_VVSlike_EquiRect 5 0.1 ./2021_MPP_SSD_alignment_EquiRect_media/images_subdiv5/ ./2021_MPP_SSD_alignment_EquiRect_media/depthmaps_full/ 0 67 67 1 ./2021_MPP_SSD_alignment_EquiRect_media/images_subdiv5/maskFull_subdiv5.png 1 0 0 1

that are, in the reading order:

  • the number of subdivision levels for the spherical image sampling
  • the Gaussian expansion parameter
  • the directory containing the images to process
  • the directory containing the depthmaps for each image
  • the reference image index (in the lexicographical order)
  • the first image index of the sequence to process
  • the last image index
  • the image sequence looping step
  • the image file of the mask (white pixels are to be considered whereas black pixels are not)
  • the number of tested initial guesses for the optimization (the one leading to the lower MPP-SSD is kept)
  • selects which estimation type to consider between 1 pure gyro, 2 incremental gyro, 3 incremental fyro with key images
  • if 1, outputs the rotation compensated equirectangular image in jpg file
  • if 1, considers truncated Gaussian domain (+ or - 3 lambda_g at most)
  • ficPosesInit the text file of initial poses, one pose line per image to process (no example provided)

Sample images, depthmaps and mask files are provided in the "MPP_SSD_alignment_EquiRect_media" archive to be downloaded from here: http://mis.u-picardie.fr/~g-caron/data/PeR/2021_MPP_SSD_alignment_EquiRect_media.zip

Copyright (C) 2021-2025 by MIS lab (UPJV). All rights reserved.

See http://mis.u-picardie.fr/~g-caron/fr/index.php?page=7 for more information.

This software was developed at:
CNRS - AIST JRL (Joint Robotics Laboratory)
1-1-1 Umezono, Tsukuba, Ibaraki
Japan

This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

Description:
Insight about how to set the project and build the program
Author:
Guillaume CARON

About

Algorithms to align images with 6 DOF

Resources

License

Stars

Watchers

Forks

Packages

No packages published