Leakage-Aware Ablation of Data Augmentation for Sugarcane Leaf Disease Classification: Operator- and Severity-Specific Effects
Abstract
Conventional image augmentation is widely used to improve model generalization; however, its benefit in fine-grained plant disease classification can be overstated when duplicated or near-duplicated images cross the train–test boundary. This study tests whether conventional augmentation remains beneficial for sugarcane leaf disease classification after controlling duplicate leakage. The original Sugarcane Leaf Dataset contained 6,748 images and was audited using MD5, perceptual hash (pHash), and structural similarity index measure (SSIM), yielding 5,389 curated images. A leakage-aware stratified 70/15/15 train–validation–test split evaluated 13 augmentation configurations using ImageNet-pretrained MobileNetV2 across three random seeds, totaling 39 runs. Macro-F1 performance was assessed through paired stratified bootstrap analysis and an exact Page test. The baseline attained the highest mean Macro-F1 (0.9646 ± 0.0068), while photometric augmentation produced comparable performance (0.9643 ± 0.0079). Geometric policies generally reduced performance, and combined augmentation achieved the lowest Macro-F1 (0.8029). The Page test indicated an overall negative ordered trend, although the medium recipe outperformed the light recipe. These results demonstrate that augmentation policies must be validated by operator and severity level, rather than enabled by default in plant disease recognition.
Related
- An innovative framework based on GMV-ALFF multi-modal MRI and Gaussian data augmentation for psychiatric disorder classification
- SeverYX: An ensemble framework for robust variant severity classification
- Explainable plant disease classification and severity assessment using MobileNetV3-large for crop health monitoring
- Fine-grained sugarcane varietal classification using high-resolution PRSS-1 satellite imagery: a field-validated dataset and a spatial context-aware deep learning framework
- Conditional spatial classification of expert-confirmed interictal epileptiform discharge epochs: An EEG-ECG ablation and SHAP analysis
BibTeX
@article{budiasto2026leakageaware,
title = {{Leakage-Aware Ablation of Data Augmentation for Sugarcane Leaf Disease Classification: Operator- and Severity-Specific Effects}},
author = {Jarot Budiasto and Hasanudin Jayawardana and Tri Kustanti Rahayu and Yosefina Mangera},
journal = {Journal of Information Systems and Informatics},
year = {2026},
month = oct,
doi = {10.63158/journalisi.v8i5.1896},
url = {https://doi.org/10.63158/journalisi.v8i5.1896},
note = {CC BY},
}
RIS
TY - JOUR
TI - Leakage-Aware Ablation of Data Augmentation for Sugarcane Leaf Disease Classification: Operator- and Severity-Specific Effects
AU - Jarot Budiasto
AU - Hasanudin Jayawardana
AU - Tri Kustanti Rahayu
AU - Yosefina Mangera
PY - 2026
DA - 2026/10/24
JO - Journal of Information Systems and Informatics
DO - 10.63158/journalisi.v8i5.1896
UR - https://doi.org/10.63158/journalisi.v8i5.1896
AB - Conventional image augmentation is widely used to improve model generalization; however, its benefit in fine-grained plant disease classification can be overstated when duplicated or near-duplicated images cross the train–test boundary. This study tests whether conventional augmentation remains beneficial for sugarcane leaf disease classification after controlling duplicate leakage. The original Sugarcane Leaf Dataset contained 6,748 images and was audited using MD5, perceptual hash (pHash), and structural similarity index measure (SSIM), yielding 5,389 curated images. A leakage-aware stratified 70/15/15 train–validation–test split evaluated 13 augmentation configurations using ImageNet-pretrained MobileNetV2 across three random seeds, totaling 39 runs. Macro-F1 performance was assessed through paired stratified bootstrap analysis and an exact Page test. The baseline attained the highest mean Macro-F1 (0.9646 ± 0.0068), while photometric augmentation produced comparable performance (0.9643 ± 0.0079). Geometric policies generally reduced performance, and combined augmentation achieved the lowest Macro-F1 (0.8029). The Page test indicated an overall negative ordered trend, although the medium recipe outperformed the light recipe. These results demonstrate that augmentation policies must be validated by operator and severity level, rather than enabled by default in plant disease recognition.
SN - 2656-5935
SN - 2656-4882
ER -