Steps to Tree Segementation Code:
- Python 3.6 version recommended
- Install
requirements.txt
bypip install -r requirements.txt
. - Navigate to the
tree segemenation folder
. - Run
train.py
- To visualize the segmentation from the pretrained model at an inference stage , run the
inference_script.ipynb
.
Steps to run the tree species classification:
- Python 3.6 version recommended
- Install
requirements.txt
bypip install -r requirements.txt
. - Navigate to the
tree species classifcation folder
. - Unzip the files in
data
folder and save them in the same nametrain
andtest
. - Ensure that
train
andtest
are the folder names. - Run the
.ipynb
file cell by cell to see the results
Steps to run planning:
- Python libraries os, numpy, scipy, pandas, seaborn, matplotlib, and pickle required. Version must be recent. Note: seaborn and pandas must be compabtible with each other.
- Navigate to
planning/src
. - Run
main.py
to run experiments. This will createplanning/outputs
and will fill with .png maps and .pkl data files. - Run
process_results.py
to process the results. This saves filesplanning/outputs/all_data_${i}.pkl
with processed data in the form of numpy arrays. - To unzip and process data, run
make_plots.py
. This file reads outputs from step 4, and creates one large dataframe. Results can be sorted, sifted, etc as desired by the user. The output plots for the current file are presented in our paper.
planning/inputs
contains the Google Maps inputs for the planning part of the project, as well as the 200ft-to-pixels ratio.
Addtionally, we also tried classifying tree species using the canopy images from Sierra Nevada forest
. This is a work in progress and still needs further investigation which is currently not under the present scopr of the project.
- Python 3.6 version recommended
- Install
requirements.txt
bypip install -r requirements.txt
. - Navigate to the
Tree_Classification_Initial_Results
and run theTraining and Testing.ipynb
.
Steps to collect data with the Sensor Logic Inc's uwb radar:
- In a windows machine, connect the radar.
- Identify the usb port (go to Device Manager and check under USB Connector Managers)
- Change the port in line 5 of
collect_data.m
- Run
collect_data.m
- Run
analyze_data.m