READ ME File For 'Dataset for Semantic segmentation of pollen grain images generated from scattering patterns via deep learning'

Dataset DOI: 10.5258/SOTON/D1741

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

This dataset supports the publication:
AUTHORS:James A. Grant-Jacob, Matthew Praeger, Robert W. Eason and Ben Mills
TITLE:Semantic segmentation of pollen grain images generated from scattering patterns via deep learning
JOURNAL: IOP Journal of Physics Communications
PAPER DOI IF KNOWN: 


This dataset contains:

Fig1.jpeg
Fig2.jpeg
Fig3.jpeg
Fig4.jpeg
Fig5.jpeg
Fig6.jpeg
Fig7.jpeg
Fig8.jpeg

Fig8.txt Data for training loss for isolated species and mixed species segmentation training.
Table1.txt Data for accuracy of the segmented images shown in Fig5.
Table2.txt Data for accuracy of the segmented images shown in Fig6.

The figures are as follows:

Fig. 1 Two-step concept showing image generation from experimental scattering patterns followed by identification via semantic segmentation of the generated images. A neural network is trained using microscope images and scattering patterns to be able to generate a microscope equivalent image of previously unseen pollen grains. Then (using only microscope images for training) a neural network is trained to segment a generated image into its constituent parts, for example as shown here, background, Iva xanthiifolia or Narcissus pollen grains.

Fig. 2 Diagram showing the experimental setup, which includes three collinear laser beams that were focussed onto pollen grains present on a glass slide. The light scattered forward from the pollen grains was collected by a camera sensor placed ~ 3 mm away from the glass slide. The pollen grains were simultaneously imaged via back-reflection using a beam splitter that reflected light from the glass slides surface onto a camera’s sensor. Inset: Iva xanthiifolia experimental scattering pattern and corresponding microscope image.

Fig. 3  Examples of the experimental scattering pattern (top), microscope image (middle) and corresponding manually labelled segmented image (bottom) of a pollen grain from species (a) Bellis perennis, (b) Populus deltoides, (c) Narcissus, (d) Iva xanthiifolia, (e) Populus tremuloides, (f) Hyacinthus orientalis, (g) Chrysanthemum, (h) Antirrhinum majus, (i) Chamelaucium and (j) Rosa. The background is labelled black.

Fig. 4  Example of augmenting training data (both microscope image and segmented colour image), showing the fusing of two separate microscope images of pollen grains of species Narcissus and Iva xanthiifolia, which had been translated prior to fusing.

Fig. 5  The capability of the neural networks for two different species of pollen grain is displayed showing, from left to right, the microscope image, generated image, the ground truth label, the predicted label and the error between the truth and the predicted label for (a) Narcissus and (b) Chamelaucium. Black in the error images corresponds to correctly labelled pixels, and white incorrectly labelled pixels.

Fig. 6  The capability of the neural networks for two different mixtures of pollen grains is shown in (a) and (b), showing, from left to right, the microscope image, generated image, the ground truth label, the predicted label and the error between the truth and the predicted label. Incorrect pixels in the error image are white and the correct pixels are black.

Fig. 7  The capability of testing the neural networks on additional pollen grain data, showing the scattering patterns and the corresponding generated images with the predicted label overlaid. Inset: the mean label accuracy and estimated size (area in µm2) of each pollen label.

Fig. 8  The capability of testing the neural networks on additional pollen grain data, showing the scattering patterns and the corresponding generated images with the predicted label overlaid. Inset: the mean label accuracy and estimated size (area in µm2) of each pollen label.

Date of data collection: 07/10/2020

Information about geographic location of data collection: Building 46, University of Southampton, Southampton, SO17 1BJ
 
Licence: CC-BY

Related projects:
EPSRC grant EP/N03368X/1
EPSRC grant EP/T026197/1


Date that the file was created: 04, 2021