Geotechnical characterisation of coal spoil piles using high-resolution optical and multispectral data - A machine learning approach

Aug 2, 2026

Geotechnical characterisation of coal spoil piles traditionally depends on field experts, which is hazardous and time-intensive. A machine-learning workflow is proposed to address the difficulty of accurately segmenting and classifying very high-resolution heterogeneous mining dump terrains using UAV remote sensing. The approach combines morphology-based segmentation with spectral, textural, structural and statistical feature extraction, applies mRMR feature selection, and performs supervised classification. The resulting automated characterisation supports proactive dump stability assessment and improves stability models while enabling more responsible mining practices.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages20

Reading Time50 min

AccessFree to download

Introduction

Environmental and safety relevance of coal spoil characterisation

Challenges of irregular deposition and property variability

Sensor selection for spectral and spatial resolution

Using object-based image analysis (OBIA) for rugged terrains

Engineering Geology 329 (2024) 107406 Contents lists available at ScienceDirect Engineering Geology journal [homepage: www.elsevier.com/locate/enggeo](homepage: www.elsevier.com/locate/enggeo) | Geotechnical characterisation of coal spoil piles using high-resolution optical and multispectral data: A machine learning approach | | | | | --- | --- | --- | --- | | Sureka Thiruchittampalama, b, Bikram Pratap Banerjee c, Nancy F. Glennd, Simit Ravala, *
a School of Minerals and Energy Resources Engineering, University of New South Wales, Sydney, NSW 2052, Australia b Department of Earth Resources Engineering, University of Moratuwa, Moratuwa 01400, Sri Lanka
c School of Surveying and Built Environment, University of Southern Queensland, Toowoomba, QLD 4350, Australia d Department of Geosciences, Boise State University, Boise, ID, USA | | | | | A R T I C L E I N F O | | A B S T R A C T | | | Keywords:
Object-based image analysis Morphology-based segmentation Waste materials
Mine dump
High-resolution UAV images Shear strength parameters | | Geotechnical characterisation of spoil piles has traditionally relied on the expertise of field specialists, which can be both hazardous and time-consuming. Although unmanned aerial vehicles (UAV) show promise as a remote sensing tool in various applications; accurately segmenting and classifying very high-resolution remote sensing images of heterogeneous terrains, such as mining spoil piles with irregular morphologies, presents significant challenges. The proposed method adopts a robust approach that combines morphology-based segmentation, as well as spectral, textural, structural, and statistical feature extraction techniques to overcome the difficulties associated with spoil pile characterisation. Additionally, it incorporates minimum redundancy maximum relevance (mRMR) based feature selection and machine learning-based classification. This automated characterisation will serve as a proactive tool for dump stability assessment, providing crucial data for improved stability models and contributing to a greener and more responsible mining industry . | | 1. Introduction Coal mining, including the construction of coal spoil dumps, poses significant environmental and safety challenges when not managed properly. The geotechnical characterisation of coal spoil is essential for monitoring and evaluating environmental impacts and safety concerns, as well as guiding effective land restoration efforts (Zevgolis et al., 2021). This data also supports policymaking, raises awareness about coal mining effects, and facilitates informed decision-making. However, the irregular deposition processes of spoil and inadequate compaction during placement can introduce complexities and uncertainties in spoil dump behavior (Masoudian et al., 2019). Analysing the spatial arrangement of characterised coal spoil is crucial, but it’s challenging due to variations in properties between mines, influenced by factors like ore type and mining techniques (Lottermoser, 2007). Traditional manual field methods for spoil characterisation are time-consuming and hazardous. To address these issues, unmanned aerial vehicles (UAVs) provide a promising solution for characterising mine materials (Yang et al., 2023). Selecting the right sensor for a specific task in mining requires considering spectral and spatial resolution. For visual inspection, mapping, and providing useful visual information, red-green-blue (RGB) sensors are commonly used (Sinaice et al., 2022; Yang et al., 2023). Multispectral sensors with expanded spectral coverage (beyond RGB) are relevant for a range of mining applications and more advanced image analysis. In addition, the calibration of multispectral images captured helps mitigate the challenges posed by variations in reflectance caused by topographic changes. In the context of spoil characterisation, selecting appropriate sensors involves understanding the challenges asso

FAQs

Why is automated geotechnical characterisation of coal spoil piles needed?

Traditional field-based characterisation is time-consuming and hazardous. Automated methods are needed to support safer, faster monitoring and assessment for stability and environmental management.

What data and segmentation strategy does the proposed method use?

The method uses high-resolution UAV optical and multispectral data. It performs morphology-based segmentation and extracts spectral, textural, structural and statistical features to handle irregular terrain.

How does the workflow improve model performance?

It applies mRMR-based feature selection to reduce redundancy and keep informative variables, then uses machine learning classification to produce characterisation results for stability assessment.

Aug 2, 2026

Upload by River Wang