Katarzyna Jabłońska
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Katarzyna Jabłońska has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Maciej Zięba
Wrocław University of Science and Technology
5 shared publications - Dariusz Tanajewski
TerraEye, Wrocaw, Poland
3 shared publications - Marcin Jakubowicz
TerraEye, Wrocaw, Poland
2 shared publications - Dawid Kopeć
WUST
1 shared publication - Marcin Maksymowicz
Remote Sensing Environmental Solutions, Jana Długosza 60A, 51-162 Wrocław, Poland
1 shared publication - Marek Wilgucki
Remote Sensing Business Solutions, Jana Długosza 60A, 51-162 Wrocław, Poland
1 shared publication
Latest Publications
- 28 June 2026
- Published by Springer Nature
- p. 504-518
Mind the Gap: Quantifying the Domain Gap in Cross-Sensor Diffusion Super-Resolution
- 28 June 2026
- Published by Springer Nature
- p. 504-518
Mind the Gap: Quantifying the Domain Gap in Cross-Sensor Diffusion Super-Resolution
- 26 June 2026
- Published by Springer Nature
- p. 19-32
Metric Learning Approach to Mineral Mapping Using Multi-sensor Satellite Data
- 26 June 2026
- Published by Springer Nature
- p. 19-32
Metric Learning Approach to Mineral Mapping Using Multi-sensor Satellite Data
- Open Access
- 21 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23 (1545598X) , 2503905
Uncertainty-Weighted Spectral Similarity for Hyperspectral Imaging Spectroscopy
Spectral similarity measures such as the Spectral Angle Mapper (SAM) treat all spectral bands equally, an assumption difficult to justify for modern imaging spectroscopy data where measurement uncertainty varies strongly with wavelength. We introduce wSAM, an uncertainty-weighted extension of SAM that incorporates per-band observational uncertainty as inverse-variance weights. On 47 EMIT–RadCalNet observation pairs across five sites, wSAM (w = 1/σ2) reduces mean spectral angle by 49% relative to SAM and outperforms both hard band masking and alternative weighting schemes, while a cross-site discriminability analysis confirms that weighting sharpens rather than erases spectral differences between surface types.
- 21 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23 (1545598X) , 2503905
Open AccessUncertainty-Weighted Spectral Similarity for Hyperspectral Imaging Spectroscopy
Spectral similarity measures such as the Spectral Angle Mapper (SAM) treat all spectral bands equally, an assumption difficult to justify for modern imaging spectroscopy data where measurement uncertainty varies strongly with wavelength. We introduce wSAM, an uncertainty-weighted extension of SAM that incorporates per-band observational uncertainty as inverse-variance weights. On 47 EMIT–RadCalNet observation pairs across five sites, wSAM (w = 1/σ2) reduces mean spectral angle by 49% relative to SAM and outperforms both hard band masking and alternative weighting schemes, while a cross-site discriminability analysis confirms that weighting sharpens rather than erases spectral differences between surface types.
- Open Access
- 11 December 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 304
Metric learning networks for accurate mineral detection with different augmentation techniques
- 11 December 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 304
Open AccessMetric learning networks for accurate mineral detection with different augmentation techniques
- 17 April 2025
- Published by Springer Nature
- p. 98-112
Satellite Data Based One-Class Classifier Models with Augmentation for Efficient Mineral Exploration
- 17 April 2025
- Published by Springer Nature
- p. 98-112
Satellite Data Based One-Class Classifier Models with Augmentation for Efficient Mineral Exploration
- Open Access
- 8 March 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (6) , 955
MineCam: Application of Combined Remote Sensing and Machine Learning for Segmentation and Change Detection of Mining Areas Enabling Multi-Purpose Monitoring
Our study addresses the need for universal monitoring solutions given the diverse environmental impacts of surface mining operations. We present a solution combining remote sensing and machine learning techniques, utilizing a dataset of over 2000 satellite images annotated with ten distinct labels indicating mining area components. We tested various approaches to develop comprehensive yet universal machine learning models for mining area segmentation. This involved considering different types of mines, raw materials, and geographical locations. We evaluated multiple satellite data set combinations to determine optimal outcomes. The results suggest that radar and multispectral data fusion did not significantly improve the models’ performance, and the addition of further channels led to the degradation of the metrics. Despite variations in mine type or extracted material, the models’ effectiveness remained within an Intersection over Union value range of 0.65–0.75. Further, in this research, we conducted a detailed visual analysis of the models’ outcomes to identify areas requiring additional attention, contributing to the discourse on effective mining area monitoring and management methodologies. The visual examination of models’ outputs provides insights for future model enhancement and highlights unique segmentation challenges within mining areas.
- 8 March 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (6) , 955
Open AccessMineCam: Application of Combined Remote Sensing and Machine Learning for Segmentation and Change Detection of Mining Areas Enabling Multi-Purpose Monitoring
Our study addresses the need for universal monitoring solutions given the diverse environmental impacts of surface mining operations. We present a solution combining remote sensing and machine learning techniques, utilizing a dataset of over 2000 satellite images annotated with ten distinct labels indicating mining area components. We tested various approaches to develop comprehensive yet universal machine learning models for mining area segmentation. This involved considering different types of mines, raw materials, and geographical locations. We evaluated multiple satellite data set combinations to determine optimal outcomes. The results suggest that radar and multispectral data fusion did not significantly improve the models’ performance, and the addition of further channels led to the degradation of the metrics. Despite variations in mine type or extracted material, the models’ effectiveness remained within an Intersection over Union value range of 0.65–0.75. Further, in this research, we conducted a detailed visual analysis of the models’ outcomes to identify areas requiring additional attention, contributing to the discourse on effective mining area monitoring and management methodologies. The visual examination of models’ outputs provides insights for future model enhancement and highlights unique segmentation challenges within mining areas.