READ ME File For 'Dataset supporting the publciation"Imaging pollen using a Raspberry Pi and LED with deep learning"'

Dataset DOI: https://doi.org/10.5258/SOTON/D3107

ReadMe Author: James A Grant-Jacob  University of Southampton https://orcid.org/0000-0002-4270-4247

This dataset supports the publication:
AUTHORS: Ben Mills, Michalis N. Zervas and James A. Grant-Jacob
TITLE: Imaging pollen using a Raspberry Pi and LED with deep learning
JOURNAL: Science of the Total Environment
PAPER DOI IF KNOWN:  https://doi.org/10.1016/j.scitotenv.2024.177084


This dataset contains:

Figure_1.tif 
Figure_2.tif
Table_1.txt
Table_2.txt

The figures are as follows:

Figure_1.tif  Fig. 1. A) Diagram of experimental setup showing the imaging setup consisting of a 20× microscope objective connected to a microscope, next to the Raspberry Pi-based sensing setup. The pollen covered glass microscope slide was translated between setups using motorised XYZ stages. B) Example of a microscope image (green outline) from the imaging setup and associated scattering pattern (red outline) captured by the Pi camera from the sensing setup. c) Schematic of transforming scattering pattern (red outline) using a neural network (yellow box), trained on a Windows workstation, into a generated image of the pollen grains (blue outline).

Figure_2.tif Fig. 2. Capability of the neural network on previously unseen pollen (Narcissus Populus, Iva and Ranunculus), and on previously unseen pollen from different plant species (Tulipa, Alliium and Taraxacum). The first row shows the scattering pattern, the second row shows the generated image, the third shows the experimental image, and the fourth shows the difference, where the darker pixels indicate regions of greater error. 

The tables are as follows:

Table 1 Number of pollen grains from each plant species used in training and testing the neural network.

Table 2 SSIM, PSNR and MSE for the generated and experimental pollen images shown in Fig. 2 (the number of pollen grain images for each species is indicate in brackets).

Licence: CC-BY

Related projects:
EPSRC grant EP/T026197/1
EPSRC grant EP/W028786/1

Date that the file was created: 10, 2024