Remote sensing-based methodology for classification of plastic foils on potatoes in Skåne
(2024) In Student thesis series INES NGEK01 20241Dept of Physical Geography and Ecosystem Science
- Abstract
- The widespread use of plastic mulch in agriculture is rapidly increasing due to intensified farming practices. While plastic mulch can enhance crop yields significantly, it also introduces challenges related to waste management and environmental degradation. This study focuses on the application of Sentinel-2A imagery and the Maximum Likelihood Classifier (MLC) to classify plastic foils used in early potato cultivation in Skåne, Sweden. By utilizing remote sensing, this research develops an efficient methodology to estimate the extent of plastic mulch usage and track its changes over the past five years.
Sentinel-2 imagery, with its high temporal and spatial resolution, offers a cost-effective solution for large-scale environmental... (More) - The widespread use of plastic mulch in agriculture is rapidly increasing due to intensified farming practices. While plastic mulch can enhance crop yields significantly, it also introduces challenges related to waste management and environmental degradation. This study focuses on the application of Sentinel-2A imagery and the Maximum Likelihood Classifier (MLC) to classify plastic foils used in early potato cultivation in Skåne, Sweden. By utilizing remote sensing, this research develops an efficient methodology to estimate the extent of plastic mulch usage and track its changes over the past five years.
Sentinel-2 imagery, with its high temporal and spatial resolution, offers a cost-effective solution for large-scale environmental monitoring. This study employs a supervised, object-based classification approach, leveraging validation data from high-resolution orthophotos and drone imagery for accuracy assessment. The classification accuracy was evaluated using three segmentation configurations, with the best-performing model achieving an overall accuracy of 84.91% and a kappa value of 0.61. The results indicate a systematic overestimation of plastic mulch areas by approximately 16%, aligning with previous accuracy assessments.
The findings reveal that plastic mulch usage in Skåne varies annually, with the highest coverage recorded in 2021 being 9250 hectares and the lowest in 2023 being 5546 hectares. This study underscores the potential of Sentinel-2 data for monitoring agricultural practices and informs the management of plastic waste in agriculture. The developed methodology not only provides precise classification of plastic mulch but also contributes to sustainable agricultural practices by enabling better waste management and resource allocation. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9164136
- author
- Homes, Lara LU
- supervisor
-
- David Tenenbaum LU
- Veiko Lehsten LU
- organization
- course
- NGEK01 20241
- year
- 2024
- type
- M2 - Bachelor Degree
- subject
- keywords
- Plastic mulch, Agricultural plastics, Sentinel-2 imagery, Maximum Likelihood Classifier (MLC), Remote Sensing
- publication/series
- Student thesis series INES
- report number
- 643
- language
- English
- id
- 9164136
- date added to LUP
- 2024-06-17 12:34:38
- date last changed
- 2024-06-17 12:34:38
@misc{9164136, abstract = {{The widespread use of plastic mulch in agriculture is rapidly increasing due to intensified farming practices. While plastic mulch can enhance crop yields significantly, it also introduces challenges related to waste management and environmental degradation. This study focuses on the application of Sentinel-2A imagery and the Maximum Likelihood Classifier (MLC) to classify plastic foils used in early potato cultivation in Skåne, Sweden. By utilizing remote sensing, this research develops an efficient methodology to estimate the extent of plastic mulch usage and track its changes over the past five years. Sentinel-2 imagery, with its high temporal and spatial resolution, offers a cost-effective solution for large-scale environmental monitoring. This study employs a supervised, object-based classification approach, leveraging validation data from high-resolution orthophotos and drone imagery for accuracy assessment. The classification accuracy was evaluated using three segmentation configurations, with the best-performing model achieving an overall accuracy of 84.91% and a kappa value of 0.61. The results indicate a systematic overestimation of plastic mulch areas by approximately 16%, aligning with previous accuracy assessments. The findings reveal that plastic mulch usage in Skåne varies annually, with the highest coverage recorded in 2021 being 9250 hectares and the lowest in 2023 being 5546 hectares. This study underscores the potential of Sentinel-2 data for monitoring agricultural practices and informs the management of plastic waste in agriculture. The developed methodology not only provides precise classification of plastic mulch but also contributes to sustainable agricultural practices by enabling better waste management and resource allocation.}}, author = {{Homes, Lara}}, language = {{eng}}, note = {{Student Paper}}, series = {{Student thesis series INES}}, title = {{Remote sensing-based methodology for classification of plastic foils on potatoes in Skåne}}, year = {{2024}}, }