Virtual-Encoder-ECNDT-2023

This repository contains scripts which generate the results presented in "Virtual encoder: a two-dimension visual odometer for NDT".

0. Table of Contents

Demo video
Dataset
1. Dataset organization
Getting_started
1. Installation
Dependecies

1. Demo video

2. Dataset

The dataset required to properly reproduce the results is avaliable at https://drive.google.com/drive/folders/1IEDjHuvrFpMRdltmPRAk1uatcMgQzyiS?usp=share_link . After downloading the "data.zip" you must copy and paste to the main project directory ("Virtual-Encoder-ECNDT-2023/") and then extract it.

2.1. Dataset Organization

Each folder represents an experiment (e.g. "air_cylindrical_longest_side") which contains:

Photos shot by Raspberry Pi ground-facing camera;
Orientation of the rigid body acquired by the inertial unit in quaternion and euler angle format;

The experiments under "calibration" folder were used exclusively for calibration purpose. Data under "planar" and "cylindrical" folder were used for generating the paper results.

+---Virtual-Encoder-ECNDT-2023
|   |   data.zip
|   |   
|   \---data
|       +---calibration
|       |   +---air_cylindrical_longest_side
|       |   |       
|       |   +---air_cylindrical_shortest_side
|       |   |       
|       |   +---air_planar_longest_side
|       |   |       
|       |   +---air_planar_shortest_side
|       |   |       
|       |   +---water_cylindrical_longest_side
|       |   |       
|       |   +---water_cylindrical_shortest_side
|       |   |       
|       |   +---water_planar_longest_side
|       |   |       
|       |   \---water_planar_shortest_side
|       |   	| 
|       |           
|       +---cylindrical
|       |   +---air_closed_loop
|       |   |       
|       |   +---air_single_x
|       |   |       
|       |   +---air_single_y
|       |   |       
|       |   +---water_closed_loop
|       |   |       
|       |   +---water_single_x
|       |   |       
|       |   \---water_single_y
|       |           
|       \---planar
|           +---air_closed_loop
|           |       
|           +---air_single_x
|           |       
|           +---air_single_y
|           |       
|           +---water_closed_loop
|           |       
|           +---water_single_x
|           |       
|           \---water_single_y
|

3. Dependecies

matplotlib # for data visualization
scipy # for signal processing algorithms
numpy # for general numerical processing
Pillow # for image reading and simple image manipulation (called PIL when imported)

4.Getting Started

4.1. Installation

4.1.a Clone github repository

cd ~
git clone https://github.com/thiagokalid/Virtual-Encoder-ECNDT-2023
cd Virtual-Encoder-ECNDT-2023

4.1.b Prerequisities

You will need to build the Virtual-Encoder-ECNDT-2023 environment by following command:

thiagokalid / virtual-encoder-ecndt-2023 Goto Github PK