Bridging Classical and Modern Computer Vision: PerceptiveNet for Tree Crown Semantic Segmentation

TL;DR

PerceptiveNet is a novel deep learning model that enhances tree crown semantic segmentation by integrating Logarithmic Gabor-parameterised convolutions and extensive contextual feature extraction, outperforming existing methods across multiple datasets.

Contribution

The paper introduces PerceptiveNet, a new model combining Log-Gabor convolutions with a robust backbone, and demonstrates its superior performance and generalization in tree crown segmentation tasks.

Findings

PerceptiveNet outperforms state-of-the-art models on tree crown datasets.

The integration of Log-Gabor convolutions improves feature extraction.

The model generalizes well across different aerial scene datasets.

Abstract

The accurate semantic segmentation of tree crowns within remotely sensed data is crucial for scientific endeavours such as forest management, biodiversity studies, and carbon sequestration quantification. However, precise segmentation remains challenging due to complexities in the forest canopy, including shadows, intricate backgrounds, scale variations, and subtle spectral differences among tree species. Compared to the traditional methods, Deep Learning models improve accuracy by extracting informative and discriminative features, but often fall short in capturing the aforementioned complexities. To address these challenges, we propose PerceptiveNet, a novel model incorporating a Logarithmic Gabor-parameterised convolutional layer with trainable filter parameters, alongside a backbone that extracts salient features while capturing extensive context and spatial information through a wider receptive field. We investigate the impact of Log-Gabor, Gabor, and standard convolutional layers on semantic segmentation performance through extensive experimentation. Additionally, we conduct an ablation study to assess the contributions of individual layers and their combinations to overall model performance, and we evaluate PerceptiveNet as a backbone within a novel hybrid CNN-Transformer model. Our results outperform state-of-the-art models, demonstrating significant performance improvements on a tree crown dataset while generalising across domains, including two benchmark aerial scene semantic segmentation datasets with varying complexities.