Registration of images to unorganized 3D point clouds using contour cues

Resource Type Date
Software 2016-09-08


The following github repository contains the code developed in this project:



This project contains the tools necessary for testing the contour-based multimodal registration algorithm.


The code provided contains a Dockefile containing all the necessay dependencies to compile the project. Scripts and are also provided to build the image and run a container with the specified images. Any environment variable stating with PHD_* will be autmatically added to the container. Please configure your datasets and results folder in the script.

Provided tests

This software is shipped jointly with a demo test of the algorithm. For running this test you will need the Stanford bunny data, both an static image of the bunny and a 3-D reconstuction. This test can be found into the autorun folder. To run this test, run python autorun/ fullpath on your machine. Additionally, a test on the color gradient algorithm is also provided. To run this test, run python autorun/ gradientTestSynth.

You can find the steps done on each test in the scripts autorun/ and autorun/gradientTestSynth.

People involved

Alba Pujol PhD Candidate
Javier Ruiz Hidalgo Associate Professor
Josep R. Casas Associate Professor