Mitigating Bad Ground Truth in Supervised Machine Learning based Crop Classification: A Multi-Level Framework with Sentinel-2 Images

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: A, Sanayya, Shetty, Amoolya, Sharma, Abhijeet, Ravichandran, Venkatesh, Gosuvarapalli, Masthan Wali, Jain, Sarthak, Nanjundiah, Priyamvada, Dutta, Ujjal Kr, Sharma, Divya
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929761421361152
author A, Sanayya
Shetty, Amoolya
Sharma, Abhijeet
Ravichandran, Venkatesh
Gosuvarapalli, Masthan Wali
Jain, Sarthak
Nanjundiah, Priyamvada
Dutta, Ujjal Kr
Sharma, Divya
author_facet A, Sanayya
Shetty, Amoolya
Sharma, Abhijeet
Ravichandran, Venkatesh
Gosuvarapalli, Masthan Wali
Jain, Sarthak
Nanjundiah, Priyamvada
Dutta, Ujjal Kr
Sharma, Divya
contents In agricultural management, precise Ground Truth (GT) data is crucial for accurate Machine Learning (ML) based crop classification. Yet, issues like crop mislabeling and incorrect land identification are common. We propose a multi-level GT cleaning framework while utilizing multi-temporal Sentinel-2 data to address these issues. Specifically, this framework utilizes generating embeddings for farmland, clustering similar crop profiles, and identification of outliers indicating GT errors. We validated clusters with False Colour Composite (FCC) checks and used distance-based metrics to scale and automate this verification process. The importance of cleaning the GT data became apparent when the models were trained on the clean and unclean data. For instance, when we trained a Random Forest model with the clean GT data, we achieved upto 70\% absolute percentage points higher for the F1 score metric. This approach advances crop classification methodologies, with potential for applications towards improving loan underwriting and agricultural decision-making.
format Preprint
id arxiv_https___arxiv_org_abs_2503_11807
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mitigating Bad Ground Truth in Supervised Machine Learning based Crop Classification: A Multi-Level Framework with Sentinel-2 Images
A, Sanayya
Shetty, Amoolya
Sharma, Abhijeet
Ravichandran, Venkatesh
Gosuvarapalli, Masthan Wali
Jain, Sarthak
Nanjundiah, Priyamvada
Dutta, Ujjal Kr
Sharma, Divya
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
In agricultural management, precise Ground Truth (GT) data is crucial for accurate Machine Learning (ML) based crop classification. Yet, issues like crop mislabeling and incorrect land identification are common. We propose a multi-level GT cleaning framework while utilizing multi-temporal Sentinel-2 data to address these issues. Specifically, this framework utilizes generating embeddings for farmland, clustering similar crop profiles, and identification of outliers indicating GT errors. We validated clusters with False Colour Composite (FCC) checks and used distance-based metrics to scale and automate this verification process. The importance of cleaning the GT data became apparent when the models were trained on the clean and unclean data. For instance, when we trained a Random Forest model with the clean GT data, we achieved upto 70\% absolute percentage points higher for the F1 score metric. This approach advances crop classification methodologies, with potential for applications towards improving loan underwriting and agricultural decision-making.
title Mitigating Bad Ground Truth in Supervised Machine Learning based Crop Classification: A Multi-Level Framework with Sentinel-2 Images
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2503.11807