This repository concentrates the experiments related to the research thesis titled Robot Navigation in Orchards Using Top-View Imagery which was submitted to the Senate of the Technion - Israel Institute of Technology in May 2019.

Author: Omer Shalev

Advisor: Associate Professor Amir Degani

The experiments were tested on Ubuntu 16.04 with ROS Kinetic.

Follow the instructions below to install all dependencies:

# Create catkin workspace
mkdir -p ~/orchards_ws/src
cd ~/orchards_ws
catkin_init_workspace

# Clone dependencies
git clone https://github.com/CearLab/orchards_top_view_aided_navigation_core
git clone https://github.com/CearLab/nelder_mead
git clone https://github.com/CearLab/astar
git clone https://github.com/CearLab/navigation -b narrow-amcl-angular-range

# Clone this repository
git clone https://github.com/CearLab/orchards_top_view_aided_navigation_experiments

# Build
catkin_make

Global configuration is in the config file. Pointers to the data is configured in data_pointers (see Lavi-April for reference).

All experiments have an associated runner model under runners. Experiment configuration is done in the runner file.

Below are the runners related to the main experiments and methods presented in the work:

1. Canopies Extraction
2. Trunks Approximation and Semantic Labeling
3. Canopy-based AMCL
4. Canopy-based ICP
5. UGV Periodic State Update
6. Semantic Global Path Planning
7. Path Update