PhD_3DPerception

Repository aimed at compiling code, papers, demos etc.. related to my PhD on 3D vision and machine learning for fruit detection and shape estimation at the university of Lincoln

UKRAS : Determining shape of strawberry crops with spherical harmonics

We present our UKRAS 2020 paper in the folder called UKRAS. You can find there the PDF for the paper and some code presented as a jupyter notebook, to try and reproduce the paper. Following the jupyter notebok let you learn about how to open and process the point cloud, compute forward spherical harmonics transform, reconstruct and plot the obtained coefficients.

If you are interested by this work please contact me by email (see profile) for any colaboration, and if the paper inspires you please cite it:

@article{louedec_cielniak_2020, 
title={Determining shape of strawberry crops with spherical harmonics}, 
DOI={10.31256/mc8hl1a}, 
journal={UKRAS20 Conference: “Robots into the real world” Proceedings}, 
author={Louedec, Justin Le and Cielniak, Grzegorz}, 
year={2020}}

Computers and electronics in agriculture : 3D shape sensing and deep learning-based segmentation ofstrawberries

Code for the submission to Computers and electronics in agriculture. You will find CNN3D model, and the training + testing files and an example of dataset loading class for the simulated data. You can find the code for PointNet on my other repository : https://github.com/lelouedec/3DNetworksPytorch For SegNet you can find a pytorch model here : https://github.com/trypag/pytorch-unet-segnet

BMVC 2021 : Gaussian map predictions for 3D surfacefeature localisation and counting

In the BMVC2021 folder, you will in due time the code for the main parts of the paper published and presented this year at BMVC 2021.