MediaComem / helmert.js

JavaScript implementation of the Helmert transformation algorithm developed in Python by Sebastien Guillaume and Samuel Schwyn from INSIT at HEIG-VD.

helmert.js

JavaScript implementation of the Helmert transformation algorithm developed in Python by Sebastien Guillaume and Samuel Schwyn from INSIT at HEIG-VD. The goal of this project is to eventually integrate it in Biodivar, Julien Mercier doctoral thesis project whithin the Media Engineering institute, also at HEIG-VD.

Warning

As of right now, this project is not architecturally sound Right now, this stuff is built on Node, but the aim is to eventually have it running in the browser.

Get it running

To run all of the scripts, you will need:

• node.js
• Python 3

git clone https://github.com/MediaComem/helmert.js.git && cd helmert.js
npm install 
pip install matplotlib

Scripts

npm run dev: Runs a demo of the JavaScript implementation using the traj.xyz files in data/

npm run py: Runs a demo of the Python implementation using those same files.

Test Protocol

The main goal of this project thus far is to try to replicate the results from the python program in the javascript implementation. To do so, we have implemented a test which does the following:

• Pseudo-ramdomly generates local and global datapoints in Python and outputs them in .xyz files in data/generated_data/testLocalData.xyz and data/generated_data/testLocalData.xyz
• The files are then fed to the Python algo and the resulting transformation is output to another file data/generated_data/result.xyz
• A Jest test compares the output of the JavaScript algorithm to the results in that file. You can run the test with the following command:

npm test