Illegal Waste Detection in Remote Sensing Images: A Case Study

2025-02-10

AI Summary

To address the challenge of monitoring illegal waste dumping, this study develops an end-to-end high-resolution remote sensing image detection framework tailored for environmental law enforcement. We propose a deep learning classification model specifically designed for Very-High-Resolution (VHR) imagery, optimizing feature representation and training strategies through multi-factor ablation studies. Crucially, we conduct the first real-world, closed-loop operational validation in collaboration with local environmental protection authorities, systematically evaluating cross-regional generalization capability. Experimental results demonstrate that the model significantly reduces expert visual interpretation time by 72% on average and successfully detects illegal dumping sites in geographically distinct regions not included in training—validating its strong transferability and operational feasibility. The core contributions lie in: (1) domain-driven model design aligned with practical enforcement requirements; (2) a novel closed-loop validation mechanism grounded in authentic law enforcement workflows; and (3) empirical evidence of robust cross-domain generalization.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Environmental crime currently represents the third largest criminal activity worldwide while threatening ecosystems as well as human health. Among the crimes related to this activity, improper waste management can nowadays be countered more easily thanks to the increasing availability and decreasing cost of Very-High-Resolution Remote Sensing images, which enable semi-automatic territory scanning in search of illegal landfills. This paper proposes a pipeline, developed in collaboration with professionals from a local environmental agency, for detecting candidate illegal dumping sites leveraging a classifier of Remote Sensing images. To identify the best configuration for such classifier, an extensive set of experiments was conducted and the impact of diverse image characteristics and training settings was thoroughly analyzed. The local environmental agency was then involved in an experimental exercise where outputs from the developed classifier were integrated in the experts' everyday work, resulting in time savings with respect to manual photo-interpretation. The classifier was eventually run with valuable results on a location outside of the training area, highlighting potential for cross-border applicability of the proposed pipeline.

Problem

• Illegal waste detection using remote sensing
• Optimizing classifier for landfill identification
• Cross-border applicability of detection pipeline

Innovation

• Remote Sensing image classification
• Semi-automatic territory scanning
• Cross-border illegal waste detection