Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MineNetCD: A Benchmark for Global Mining Change Detection on Remote Sensing Imagery

Weikang Yu, Xiaokang Zhang, Xiao Xiang Zhu, Richard Gloaguen, Pedram Ghamisi

TL;DR

MineNetCD tackles the challenge of globally monitoring mining-driven land changes from remote sensing by introducing a comprehensive benchmark comprising a large-scale bi-temporal dataset, a change-aware frequency-domain baseline (ChangeFFT), and a unified framework (UCD) that democratizes access to multiple change-detection models via HuggingFace. The ChangeFFT module learns change representations in the frequency domain to align bi-temporal feature spectra, while a Swin-Transformer/ResNet/VMamba-enabled modular Siamese encoder provides flexible backbones. Empirical results show MineNetCD outperforms 12 state-of-the-art methods on the benchmark, with notable gains in F1 and cIoU, and ablations confirm the efficacy of ChangeFFT across backbones. By offering open dataset access and pretrained models, MineNetCD facilitates reproducibility and accelerates research toward environmentally responsible mining monitoring using remote sensing.

Abstract

Monitoring changes triggered by mining activities is crucial for industrial controlling, environmental management and regulatory compliance, yet it poses significant challenges due to the vast and often remote locations of mining sites. Remote sensing technologies have increasingly become indispensable to detect and analyze these changes over time. We thus introduce MineNetCD, a comprehensive benchmark designed for global mining change detection using remote sensing imagery. The benchmark comprises three key contributions. First, we establish a global mining change detection dataset featuring more than 70k paired patches of bi-temporal high-resolution remote sensing images and pixel-level annotations from 100 mining sites worldwide. Second, we develop a novel baseline model based on a change-aware Fast Fourier Transform (ChangeFFT) module, which enhances various backbones by leveraging essential spectrum components within features in the frequency domain and capturing the channel-wise correlation of bi-temporal feature differences to learn change-aware representations. Third, we construct a unified change detection (UCD) framework that integrates over 13 advanced change detection models. This framework is designed for streamlined and efficient processing, utilizing the cloud platform hosted by HuggingFace. Extensive experiments have been conducted to demonstrate the superiority of the proposed baseline model compared with 12 state-of-the-art change detection approaches. Empirical studies on modularized backbones comprehensively confirm the efficacy of different representation learners on change detection. This contribution represents significant advancements in the field of remote sensing and change detection, providing a robust resource for future research and applications in global mining monitoring. Dataset and Codes are available via the link.

MineNetCD: A Benchmark for Global Mining Change Detection on Remote Sensing Imagery

TL;DR

MineNetCD tackles the challenge of globally monitoring mining-driven land changes from remote sensing by introducing a comprehensive benchmark comprising a large-scale bi-temporal dataset, a change-aware frequency-domain baseline (ChangeFFT), and a unified framework (UCD) that democratizes access to multiple change-detection models via HuggingFace. The ChangeFFT module learns change representations in the frequency domain to align bi-temporal feature spectra, while a Swin-Transformer/ResNet/VMamba-enabled modular Siamese encoder provides flexible backbones. Empirical results show MineNetCD outperforms 12 state-of-the-art methods on the benchmark, with notable gains in F1 and cIoU, and ablations confirm the efficacy of ChangeFFT across backbones. By offering open dataset access and pretrained models, MineNetCD facilitates reproducibility and accelerates research toward environmentally responsible mining monitoring using remote sensing.

Abstract

Monitoring changes triggered by mining activities is crucial for industrial controlling, environmental management and regulatory compliance, yet it poses significant challenges due to the vast and often remote locations of mining sites. Remote sensing technologies have increasingly become indispensable to detect and analyze these changes over time. We thus introduce MineNetCD, a comprehensive benchmark designed for global mining change detection using remote sensing imagery. The benchmark comprises three key contributions. First, we establish a global mining change detection dataset featuring more than 70k paired patches of bi-temporal high-resolution remote sensing images and pixel-level annotations from 100 mining sites worldwide. Second, we develop a novel baseline model based on a change-aware Fast Fourier Transform (ChangeFFT) module, which enhances various backbones by leveraging essential spectrum components within features in the frequency domain and capturing the channel-wise correlation of bi-temporal feature differences to learn change-aware representations. Third, we construct a unified change detection (UCD) framework that integrates over 13 advanced change detection models. This framework is designed for streamlined and efficient processing, utilizing the cloud platform hosted by HuggingFace. Extensive experiments have been conducted to demonstrate the superiority of the proposed baseline model compared with 12 state-of-the-art change detection approaches. Empirical studies on modularized backbones comprehensively confirm the efficacy of different representation learners on change detection. This contribution represents significant advancements in the field of remote sensing and change detection, providing a robust resource for future research and applications in global mining monitoring. Dataset and Codes are available via the link.
Paper Structure (19 sections, 7 equations, 15 figures, 3 tables)

This paper contains 19 sections, 7 equations, 15 figures, 3 tables.

Table of Contents

  1. Introduction
  2. MineNetCD Dataset
  3. MineNetCD Model
  4. Modularized Siamese Encoder
  5. Change-aware Fast Fourier Transform (ChangeFFT)
  6. Change Decoder
  7. Loss Function
  8. Unified Change Detection Framework
  9. Experiments
  10. Dataset Preparation
  11. Baseline Approaches
  12. Experimental Settings
  13. Evaluation Metrics
  14. Benchmark Results with the State-of-the-art
  15. Quantitative Results
  16. ...and 4 more sections

Figures (15)

  • Figure 1: The MineNetCD.
  • Figure 2: Spatial distribution of the mining sites in MineNetCD dataset
  • Figure 3: To locate the mining sites: (a) Overlay the global mining polygon data maus2020global onto Google Earth. This allows the sites to be accurately identified. (b) and (c) illustrate two such examples.
  • Figure 4: Typical samples in MineNetCD. (a) and (b) Pre- and post-event imagery. (c) Change masks. (d) Spectral intensity histograms.
  • Figure 5: Statistics of MineNetCD. (a) No. of mining sites in each country. Changed areas of (b) each country and (c) each site. (d) Image acquisition time.
  • ...and 10 more figures