What question did this study set out to answer?

The aim is to assess the accuracy of UAV-derived RGB imagery for detecting wild boar damage in rapeseed fields.

March 1, 2026Open Access

Deep Learning Segmentation Models for UAV-Based Detection of Crop Damage in Rapeseed Using RGB Imagery

Puntos clave

The aim is to assess the accuracy of UAV-derived RGB imagery for detecting wild boar damage in rapeseed fields.
Utilized four deep learning architectures: U-Net, U-Net++, DeepLabV3+, and PSPNet.
Conducted experiments on rapeseed fields in central-western Poland using UAV imagery.
Assessed model performance with overall accuracy, F1-score, and Intersection over Union (IoU).
Tested input configurations using RGB imagery alone and with topographic position index (TPI) derived from a digital surface model.
Achieved overall accuracy for RGB-only data between 0.957 and 0.972.
Best-performing model (U-Net) had a damaged-class F1 score of 0.752 and IoU of 0.603.
Supplementing RGB data with TPI showed minimal change in overall accuracy and damaged-class metrics.
Non-damaged areas were classified very accurately with F1 scores > 0.977 and IoU > 0.955.

Resumen

The objective of this study was to evaluate the accuracy of detecting crop damage caused by wild boar in rapeseed fields using UAV (unmanned aerial vehicle)-derived RGB (red, green and blue) imagery and deep learning segmentation models. The experiments were conducted on rapeseed crops at full maturity shortly before harvest in central-western Poland in 2021. Four convolutional neural network architectures—U-Net (U-shaped network), U-Net++, DeepLabV3+ (deep learning + labelling), and PSPNet (Pyramid Scene Parsing Network)—were benchmarked using two input configurations: RGB imagery alone and RGB combined with the topographic position index (TPI) derived from a digital surface model (DSM). Model performance was assessed using overall accuracy, F1-score (harmonic mean of precision and recall), and Intersection over Union (IoU), with class-specific metrics reported to provide a realistic evaluation of damaged-area detection. For RGB-only data, overall accuracy ranged from 0.957 to 0.972, while damaged-class F1 and IoU reached 0.752 and 0.603, respectively, for the best-performing model (U-Net). When RGB data were supplemented with TPI, overall accuracy and damaged-class metrics changed only slightly, indicating limited benefit from the topographic feature under these field conditions. Non-damaged crop areas were consistently well-classified (F1 > 0.977, IoU > 0.955). These results confirm that UAV-based RGB imagery enables reliable late-season assessment of wildlife-induced crop damage, and that reporting class-specific metrics in spatially independent test sets is essential for realistic performance evaluation.

Leer artículo completoexternamente

Me gusta

Guardar

Ver artículo completo