A Hybrid Random Forest and CNN Framework for Tile-Wise Oil-Water Classification in Hyperspectral Images
Mehdi Nickzamir, Seyed Mohammad Sheikh Ahamdi Gandab
TL;DR
The study tackles oil-water classification in hyperspectral images under severe class imbalance by introducing a tile-wise hybrid framework that couples Random Forest with a convolutional neural network. The RF handles high dimensional spectral data while the CNN refines probability maps to capture spatial context, yielding improved recall, F1, and AUC. On the HOSD dataset, the approach achieves a recall of 0.85, F1 of 0.84, and AUC of 0.99, representing improvements over baselines by notable margins. This tile wise spectral-spatial fusion demonstrates robust and context aware oil spill detection with potential for scalable environmental monitoring.
Abstract
A novel hybrid Random Forest and Convolutional Neural Network (CNN) framework is presented for oil-water classification in hyperspectral images (HSI). To address the challenge of preserving spatial context, the images were divided into smaller, non-overlapping tiles, which served as the basis for training, validation, and testing. Random Forest demonstrated strong performance in pixel-wise classification, outperforming models such as XGBoost, Attention-Based U-Net, and HybridSN. However, Random Forest loses spatial context, limiting its ability to fully exploit the spatial relationships in hyperspectral data. To improve performance, a CNN was trained on the probability maps generated by the Random Forest, leveraging the CNN's capacity to incorporate spatial context. The hybrid approach achieved 7.6% improvement in recall (to 0.85), 2.4% improvement in F1 score (to 0.84), and 0.54% improvement in AUC (to 0.99) compared to the baseline. These results highlight the effectiveness of combining probabilistic outputs with spatial feature learning for context-aware analysis of hyperspectral images.