Centralized Copy-Paste: Enhanced Data Augmentation Strategy for Wildland Fire Semantic Segmentation

2025-07-08

AI Summary

To address the scarcity of flame annotations and low segmentation accuracy for fire classes in wildfire semantic segmentation, this paper proposes a Centralized Copy-Paste (CCP) data augmentation method. CCP precisely extracts flame regions from source images, identifies their spatial centroids via clustering, crops the most salient central subregions, and seamlessly embeds them into target images using geometry- and illumination-adaptive fusion. Crucially, CCP introduces a region-centering mechanism to enhance discriminative representation of flame-specific features, thereby significantly improving model generalization under limited-label conditions. Within a multi-class segmentation framework, CCP is evaluated using weighted loss and multi-objective optimization. Experiments demonstrate that CCP consistently outperforms state-of-the-art augmentation methods in both IoU and F1-score, achieving superior flame segmentation performance on both public and in-house wildfire datasets.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Collecting and annotating images for the purpose of training segmentation models is often cost prohibitive. In the domain of wildland fire science, this challenge is further compounded by the scarcity of reliable public datasets with labeled ground truth. This paper presents the Centralized Copy-Paste Data Augmentation (CCPDA) method, for the purpose of assisting with the training of deep-learning multiclass segmentation models, with special focus on improving segmentation outcomes for the fire-class. CCPDA has three main steps: (i) identify fire clusters in the source image, (ii) apply a centralization technique to focus on the core of the fire area, and (iii) paste the refined fire clusters onto a target image. This method increases dataset diversity while preserving the essential characteristics of the fire class. The effectiveness of this augmentation technique is demonstrated via numerical analysis and comparison against various other augmentation methods using a weighted sum-based multi-objective optimization approach. This approach helps elevate segmentation performance metrics specific to the fire class, which carries significantly more operational significance than other classes (fuel, ash, or background). Numerical performance assessment validates the efficacy of the presented CCPDA method in alleviating the difficulties associated with small, manually labeled training datasets. It also illustrates that CCPDA outperforms other augmentation strategies in the application scenario considered, particularly in improving fire-class segmentation performance.

Problem

• Addresses scarcity of labeled wildland fire datasets
• Improves fire-class segmentation in deep-learning models
• Enhances dataset diversity while preserving fire characteristics

Innovation

• Centralized Copy-Paste for fire segmentation
• Focus on core fire area characteristics
• Multi-objective optimization enhances fire-class metrics