Dataset for Accurate non-invasive quantification of astaxanthin content using hyperspectral images and machine learning

Calderini, Marco; Yli-Tuomola, Aliisa; Pölönen, Ilkka; Timilsina, Hemanta; Pulkkinen, Katja; Pääkkönen, Salli; Salmi, Pauliina. Dataset for Accurate non-invasive quantification of astaxanthin content using hyperspectral images and machine learning. V. 13.9.2024. University of Jyväskylä. https://doi.org/https://doi.org/10.23729/be85c733-2555-45b9-8bca-5fe441beb48b.

All authors: Calderini, Marco; Yli-Tuomola, Aliisa; Pölönen, Ilkka; Timilsina, Hemanta; Pulkkinen, Katja; Pääkkönen, Salli; Salmi, Pauliina

Funders: Research Council of Finland

Right-holders: 

Availability and identifiers

Availability: Direct download (Embargo date: 01/03/2025)

DOI identifier in original repository: https://doi.org/https://doi.org/10.23729/be85c733-2555-45b9-8bca-5fe441beb48b

Description of the dataset

Description: The dataset contains spectral data of cell suspensions of the microalgae Haematococcus pluvialis under no-stress and stress conditions. Spectral data was obtained with a hyperspectral imager (reflectance) and a spectrophotometer coupled with an integrating sphere (absorbance). Together with the raw data files, this dataset contains the Jupyter Notebook (PYTHON language) scripts to process the data and analysed it. Among the analysis, linear models and a convolutional neural network (CNN) are developed for the spectral data. The objective of this dataset was to develop a CNN able to accurately quantify astaxanthin content per dry weight from hyperspectral images (HSI). The CNN prediction accuracy was compared to linear models using the spectrophotometer couples with the integrating sphere. In addition to the scripts, this dataset contains all data files generated in those scripts.

Language: English

Keywords (YSO): algae; monitoring; machine learning; pigments

Fields of science: 113 Computer and information sciences

Follow-up groups: Aquatic Sciences (Department of Biological and Environmental Science BIOENV) WET; School of Resource Wisdom (University of Jyväskylä JYU) JYU.Wisdom; Computational Science (Faculty of Information Technology IT) LASK

Do you deal with data concerning special categories of personal data in your research?: No

Projects related to dataset

Publications and other outputs related to dataset