Remote Sensing Image Interpretation Research Articles

AFA–Mamba: Adaptive Feature Alignment with Global–Local Mamba for Hyperspectral and LiDAR Data Classification

The joint classification of hyperspectral imagery (HSI) and LiDAR data is an important task in the field of remote sensing image interpretation. Traditional classification methods, such as support vector machine (SVM) and random forest (RF), have difficulty capturing the complex spectral–spatial–elevation correlation information. Recently, important progress has been made in HSI-LiDAR classification using Convolutional Neural Networks (CNNs) and Transformers. However, due to the large spatial extent of remote sensing images, the vanilla Transformer and CNNs struggle to effectively capture global context. Moreover, the weak misalignment between multi-source data poses challenges for their effective fusion. In this paper, we introduce AFA–Mamba, an Adaptive Feature Alignment Network with a Global–Local Mamba design that achieves accurate land cover classification. It contains two main core designs: (1) We first propose a Global–Local Mamba encoder, which effectively models context through a 2D selective scanning mechanism while introducing local bias to enhance the spatial features of local objects. (2) We also propose an SSE Adaptive Alignment and Fusion (A2F) module to adaptively adjust the relative positions between multi-source features. This module establishes a guided subspace to accurately estimate feature-level offsets, enabling optimal fusion. As a result, our AFA–Mamba consistently outperforms state-of-the-art multi-source fusion classification approaches across multiple datasets.

Scale-wised feature enhancement network for change captioning of remote sensing images

ABSTRACT The Remote Sensing Image Change Captioning (RSICC) has recently emerged in the field of remote sensing image interpretation; it aims to automatically predict natural language captions of significant semantic changes in bi-temporal remote sensing images. Recent studies of RSICC have improved the accuracy of change captions of bi-temporal remote sensing images to a large extent. Nevertheless, there still remain challenges in multi-scale perception of ground objects and feature enhancement of bi-temporal remote sensing images. To address these challenges and further improve the accuracy of RSICC, a novel deep learning–based end-to-end scale-wised feature enhancement network (SFEN) is proposed in this paper. SFEN integrates four efficient blocks: 1) the siamese backbone network (SBN) to extract initial features of bi-temporal remote sensing images, 2) the siamese receptive field fusion (SRFF) block to explicitly capture multi-scale semantic information of ground objects in bi-temporal feature maps, 3) the siamese global feature enhancement (SGFE) block to adaptively enhance key information and filtering redundant features of bi-temporal feature maps in both channel and spatial dimensions, 4) the change caption decoder (CCD) to map bi-temporal feature maps into natural language. The SFEN aims to precisely capture significant semantic information of ground objects in bi-temporal remote sensing images and predict accurate change captions. Experimental results on LEVIR-CC dataset demonstrate our SFEN outperforms recent state-of-the-art (SOTA) approach in RSICC by 5.2% on CIDEr-D and achieves a new SOTA.

Driving mechanisms and multi-scenario simulation of land use change based on National Land Survey Data: a case in Jianghan Plain, China

The Jianghan Plain is simultaneously responsible for ecological protection, food security and urbanization, and land use conflicts are prominent. Revealing the driving mechanism of land use/cover change (LUCC) and simulating the land use pattern can help to coordinate the land use conflicts in the future. Utilizing the National Land Survey Data (NLSD) of Jiangling County in Jianghan Plain (2011–2020) and the patch-generating land use simulation (PLUS) model, this paper analyzed the characteristics of land use evolution, applied the random forest classification (RFC) to analyze the driving mechanism, simulated the 2035 land use pattern under three scenarios of natural development, planning guidance and ecological protection through Markov and Cellular Automaton based on multiple random seeds (CARS) models, and proposed several countermeasures. The study found that: 1) From 2011 to 2020, town construction land increased, village construction land, agricultural land and ecological land decreased. 2) The factors driving LUCC were socio-economic factors, spatial factors, and natural factors in descending order. 3) In the three scenarios, the trend of construction land expansion, agricultural land and ecological land encroachment is inevitable by 2035. 4) It is imperative to actively advocate for large-scale mechanization and informatization of agricultural production, encourage the repurposing of idle and inefficiently used construction land, facilitate multi-purpose land utilization, and implement a policy of locally balancing occupation and compensation for cultivated and ecological land. 5) When employing the PLUS model to simulate LUCC, using continuous NLSD yielded more accurate results than remote sensing image interpretation data. This study offers a theoretical basis for the coordinated development of land use in Jianghan Plain, and presents a method to enhance the simulation accuracy of the PLUS model.

Spatial Gradient Differences in the Cooling Island Effect and Influencing Factors of Urban Park Green Spaces in Beijing

Urban park green spaces exhibit significant cool island effects, which can effectively mitigate the urban heat environment. Clarifying the characteristics and differences in the cool island effects of urban parks across different spatial gradients within cities is instrumental in identifying potential issues and optimizing the structure and resource allocation of park green spaces in a scientifically rational manner. This study focuses on parks within the central urban area of Beijing, utilizing remote sensing image interpretation and land surface temperature (LST) inversion to extract relevant characteristics of park green spaces and the park cool island intensity (PCI) index. Various mathematical and statistical methods including correlation analysis, regression analysis, and cluster analysis are employed to conduct comparative studies across three gradients: within the 3rd ring road, between the 3rd and 5th ring roads, and outside the 5th ring road. The analysis reveals that both park green space characteristics and urban heat island effects exhibit spatial gradient differences, collectively influencing the cool island effects of urban park green spaces. PCI gradually decreases across the three spatial gradients. Cluster analysis identifies four distinct types of parks with different cool island effect characteristics, highlighting the need for optimization and improvement in over half of the parks. Various indicators of park green space characteristics show different correlations with PCI, with variations in correlation strength and thresholds across gradients. The fitting effects of regression equations for each characteristic indicator and PCI gradually worsen from within the 3rd ring road to outside the 5th ring road, with different factors playing important roles across gradients. This study enhances our understanding of the cool island effects of urban park green spaces and facilitates the proposition of differentiated optimization management strategies for urban park green space planning and system construction in different regions.

Evolution characteristics, carbon emission effects and influencing factors of production-living-ecological space in Taihang Mountain poverty belt, China

The rapid advancement of urbanization and industrialization in China has gradually spread to the poor mountainous areas, which has not only brought about rapid economic development but has also caused the increasing competition for production-living-ecological spaces (PLES) and many ecological and environmental problems, carbon emissions have also increased. As an economically less developed and ecologically fragile area in China, whether the transition of the PLES in the mountain poverty belt has unique characteristics? How the PLES transition in mountainous areas affects carbon emissions and what are the important factors affecting carbon emissions? To explore these issues in depth, we studied the Taihang Mountain area in Shijiazhuang (TMS) using remote sensing image interpretation data from 2000, 2010, and 2020, and we analyzed the PLES evolution characteristics, carbon emission changes, carbon emission effects and its influencing factors of PLES. The results are as follows: 1) The TMS was dominated by ecological and production space. From 2000 to 2020, the production space decreased by 384.66 km2, the ecological space increased by 123.80 km2, and the living space increased by 260.86 km2. Agricultural production space was mainly converted to ecological and rural living space. Industrial and mining productive space was mainly converted to agricultural productive space and urban living space. 2) The study area was in a state of carbon deficit, the transition of ecological space and agricultural productive space to industrial and mining productive space and living space were the main transition types caused the carbon emissions increasing, and that of industrial and mining productive space to agricultural productive space was the main type caused the carbon emissions decreasing. 3) The proportion of construction land, urbanization rate and proportion of secondary industry are the main factors leading to the increase of carbon emissions. Per capita energy consumption, forest coverage and proportion of tertiary industry are the main factors leading to the decrease of carbon emissions. This can provide new ideas for research on carbon emissions from land-use changes and a theoretical basis for the optimization of territorial space in the mountainous areas of China.

PMNet: a multi-branch and multi-scale semantic segmentation approach to water extraction from high-resolution remote sensing images with edge-cloud computing

In the field of remote sensing image interpretation, automatically extracting water body information from high-resolution images is a key task. However, facing the complex multi-scale features in high-resolution remote sensing images, traditional methods and basic deep convolutional neural networks are difficult to effectively capture the global spatial relationship of the target objects, resulting in incomplete, rough shape and blurred edges of the extracted water body information. Meanwhile, massive image data processing usually leads to computational resource overload and inefficiency. Fortunately, the local data processing capability of edge computing combined with the powerful computational resources of cloud centres can provide timely and efficient computation and storage for high-resolution remote sensing image segmentation. In this regard, this paper proposes PMNet, a lightweight deep learning network for edge-cloud collaboration, which utilises a pipelined multi-step aggregation method to capture image information at different scales and understand the relationships between remote pixels through horizontal and vertical spatial dimensions. Also, it adopts a combination of multiple decoding branches in the decoding stage instead of the traditional single decoding branch. The accuracy of the results is improved while reducing the consumption of system resources. The model obtained F1-score of 90.22 and 88.57 on Landsat-8 and GID remote sensing image datasets with low model complexity, which is better than other semantic segmentation models, highlighting the potential of mobile edge computing in processing massive high-resolution remote sensing image data.

Study on the Spatiotemporal Evolution and Influencing Factors of Forest Coverage Rate (FCR): A Case Study on Yunnan Province Based on Remote Sensing Image Interpretation

The study of the forest coverage rate (FCR) is related to the ecological environment and sustainable development goals (SDGs) of a region. In light of the lack of an organic integration method of “spatiotemporal evolution, correlation analysis, and change prediction” and the lack of a methodology that integrates methods of “remote sensing (RS) and GIS, multi-phase LUCC, and construction of econometric models” in the research methods at present, this study focus on Yunnan, a typical border province located in China with a relatively fragile “innate” ecological environment, as the research area. Based on the interpretation of land use/land cover (LULC) data retrieved from seven periods RS images (1990, 1995, 2000, 2005, 2010, 2015, and 2020), the spatiotemporal evolution of FCR in 129 counties was analyzed. Complementary research methods, such as the spatial econometric model, geographically weighted regression (GWR), and the geographic detector (GD), are used to reveal the influencing factors of FCR. Finally, this study predicts the FCRs of 129 counties in Yunnan from 2025 to 2050. The FCR in Yunnan presents an increasing trend year by year, increasing from 28.96% in 1990 to 49.05% in 2020. In addition, it exhibits spatial agglomeration characteristics with fewer values in the east and more in the west. The analysis of influencing factors show that the increases in the per capita GDP, land utilization rate, and annual average temperature, and the implementation of the Conversion of Cultivated Land into Forest Project (CCFP) will significantly improve the FCR, while the increases in the population density land reclamation rate, the proportion of construction land area, and the proportion of soil erosion land area will significantly reduce the FCR. Furthermore, the FCR is influenced by multiple factors, and the relative factors observed not only show significant spatial differences, but also present complex and diverse patterns, with the additional characteristics of being interwoven and overlapping. This study contributes to expanding and improving the methods and pathways of exploring the spatiotemporal evolution characteristics of FCR in ecologically fragile areas using RS methods, providing a reference for increasing FCR and improving the ecological environment’s quality in Yunnan Province and other ecologically fragile areas.

Holocene activity and seismic surface rupture zone of the Abuduo fault eastern Tibetan Plateau

The Garzê–Yushu fault is a large active NW-trending strike-slip fault on the Tibetan Plateau, along which strong earthquakes have frequently occurred historically. Together with the Xianshuihe fault, the Garzê–Yushu fault constitutes the northern boundary of the Sichuan–Yunnan block, and it is also the boundary between the Bayan Har and Qiangtang blocks. The Abuduo fault is a near E–W-trending fault on the southern side of the Garzê–Yushu fault. Through remote sensing image interpretation and field seismic geological investigations, this study found a series of left-lateral displacement landforms and well-preserved seismic surface rupture zones along the Abuduo fault, extending over a distance of approximately 65 km from Abuduo through Yushu and Qinghai to Selikou Village. According to the geological and geomorphological evidence, the Abuduo fault is considered a left-lateral strike-slip fault with Holocene activity, which has tectonic conditions suitable for breeding and generating strong earthquakes. Additionally, the eastward extension of the Abuduo fault is likely to intersect the Garzê–Yushu fault and thereby decompose its horizontal sliding deformation.

Leveraging Permuted Image Restoration for Improved Interpretation of Remote Sensing Images

Leveraging Permuted Image Restoration for Improved Interpretation of Remote Sensing Images

Panoptic Perception: A Novel Task and Fine-grained Dataset for Universal Remote Sensing Image Interpretation

Panoptic Perception: A Novel Task and Fine-grained Dataset for Universal Remote Sensing Image Interpretation

Generative ConvNet Foundation Model With Sparse Modeling and Low-Frequency Reconstruction for Remote Sensing Image Interpretation

Generative ConvNet Foundation Model With Sparse Modeling and Low-Frequency Reconstruction for Remote Sensing Image Interpretation

Study on the late quaternary activity of the Jinyang-Ningnan segment of the Lianfeng fault zone

The Lianfeng fault zone is located at the junction of the Daliangshan subblock and the South China block and is the southern boundary of the Daliangshan subblock. It is a very active tectonic area, which is of great significance to understanding its seismogenic environment. However, there are relatively little existing data on its activity and the timing of its latest activity. In the Jinyang-Ningnan segment of the Lianfeng fault zone along the Jinsha River, remote sensing image interpretation, field geologic surveys, high-precision drone measurements, and sample dating indicate that the river terraces T1 (the lowest, 15–40 m above the local river level) and T2 (60–90 m above the local river level) developed in the mid-Holocene and late Pleistocene. This indicates that the Jinsha River basin has been strongly downcutting since the Late Pleistocene. The fault profiles are mainly exposed in the T1 and T2 terraces of the Jinsha River. Based on this information in combination with the fault characteristics of the sedimentary profile and stratigraphic dating, it is believed that the Lianfeng fault zone was active at least along the Jinyang-Ningnan segment, with the latest activity occurring no earlier than the early to mid-Holocene.

U-Net-LSTM: Time Series-Enhanced Lake Boundary Prediction Model

Change detection of natural lake boundaries is one of the important tasks in remote sensing image interpretation. In an ordinary fully connected network, or CNN, the signal of neurons in each layer can only be propagated to the upper layer, and the processing of samples is independent at each moment. However, for time-series data with transferability, the learned change information needs to be recorded and utilized. To solve the above problems, we propose a lake boundary change prediction model combining U-Net and LSTM. The ensemble of LSTMs helps to improve the overall accuracy and robustness of the model by capturing the spatial and temporal nuances in the data, resulting in more precise predictions. This study selected Lake Urmia as the research area and used the annual panoramic remote sensing images from 1996 to 2014 (Lat: 37°00′ N to 38°15′ N, Lon: 46°10′ E to 44°50′ E) obtained by Google Earth Professional Edition 7.3 software as the research data set. This model uses the U-Net network to extract multi-level change features and analyze the change trend of lake boundaries. The LSTM module is introduced after U-Net to optimize the predictive model using historical data storage and forgetting as well as current input data. This method enables the model to automatically fit the trend of time series data and mine the deep information of lake boundary changes. Through experimental verification, the model’s prediction accuracy for lake boundary changes after training can reach 89.43%. Comparative experiments with the existing U-Net-STN model show that the U-Net-LSTM model used in this study has higher prediction accuracy and lower mean square error.

TTNet: A Temporal-Transform Network for Semantic Change Detection Based on Bi-Temporal Remote Sensing Images

Semantic change detection (SCD) holds a critical place in remote sensing image interpretation, as it aims to locate changing regions and identify their associated land cover classes. Presently, post-classification techniques stand as the predominant strategy for SCD due to their simplicity and efficacy. However, these methods often overlook the intricate relationships between alterations in land cover. In this paper, we argue that comprehending the interplay of changes within land cover maps holds the key to enhancing SCD’s performance. With this insight, a Temporal-Transform Module (TTM) is designed to capture change relationships across temporal dimensions. TTM selectively aggregates features across all temporal images, enhancing the unique features of each temporal image at distinct pixels. Moreover, we build a Temporal-Transform Network (TTNet) for SCD, comprising two semantic segmentation branches and a binary change detection branch. TTM is embedded into the decoder of each semantic segmentation branch, thus enabling TTNet to obtain better land cover classification results. Experimental results on the SECOND dataset show that TTNet achieves enhanced performance when compared to other benchmark methods in the SCD task. In particular, TTNet elevates mIoU accuracy by a minimum of 1.5% in the SCD task and 3.1% in the semantic segmentation task.

A Study on Spatiotemporal Changes of Ecological Vulnerability in Yunnan Province Based on Interpretation of Remote Sensing Images

The inherent ecological environment of mountainous regions is highly fragile, and the degree of sustainable development is low. There has not yet been a multi-phase ecological vulnerability evaluation (EVE) study based on remote sensing (RS) and GIS for mountainous provinces, for which there is an urgent need to establish a system that is appropriate, practicable and easily operated and applied. In this study, an integrated “RS and GIS + multi-phase land use/cover change (LUCC) + practically quantitative theory and methods of EVE” approach was adopted for analysis based on the interpretation results of five phases of the land use/land cover (LULC) RS images of Yunnan, with 129 counties being considered as the evaluation units. The organic combination of quantitative multi-index comprehensive evaluation (QMCE) and qualitative comprehensive analysis (QCA) methods was adopted to perform quantitative calculations of a system of county-level evaluation indicators which includes “innate” natural ecological vulnerability (INEV), land use ecological vulnerability (LUEV) and land cover ecological vulnerability (LCEV); the degree of ecological vulnerability (DEV) was assessed for the 129 counties within the province during the five study phases (1980, 1990, 2000, 2010 and 2020). The spatiotemporal variation characteristics and laws of DEV from 1980 to 2020 in the whole province and 129 counties were revealed, aiming to provide a basis for meeting the SDGs for mountainous provinces. The results are as follows: (1) Overall, INEV is high because of the high mountains and steep slopes, and the entire province is classified as “highly vulnerable” on average. In terms of counties, more than 79.07% are classified as “moderately vulnerable”, “highly vulnerable” and “very highly vulnerable”. (2) The degree of LUEV and LCEV caused by acquired human socioeconomic activities was higher in 1980. However, after a series of ecological measures in the past 40 years, the values of DEVLU and DEVLC in the whole province and counties in 2020 have decreased to different degrees. Accordingly, the degree of overall ecological vulnerability of Yunnan province and counties decreased significantly from 1980 to 2020. The basic law of change is that the number of counties with high DEV decreases significantly, while the number of counties with low DEV increases significantly. (3) The regional difference in the DEV of Yunnan province is large. In general, the degree of ecological vulnerability is lower in the southern, southwestern, western and central areas of Yunnan and higher in the northwest high mountain canyon, northeast mountain areas and east and southeast karst areas. (4) Overall, the DEV in Yunnan province is currently still high. There is an urgent need to enhance the construction of ecological civilization across the whole province and take effective measures to protect the ecological environment according to local conditions, so as to steadily reduce the DEV.

Oriented Object Detection in Remote Sensing Using an Enhanced Feature Pyramid Network

Object detection in remote sensing images is a critical task within the field of remote sensing image interpretation and analysis, serving as a fundamental foundation for military surveillance and traffic guidance. Recently, although many object detection algorithms have been improved to adapt to the characteristics of remote sensing images and have achieved good performance, most of them still use horizontal bounding boxes, which struggle to accurately mark targets with multiple angles and dense arrangements in remote sensing images. We propose an oriented bounding box optical remote sensing image object detection method based on an enhanced feature pyramid, and add an attention module to suppress background noise. To begin with, we incorporate an angle prediction module that accurately locates the detection target. Subsequently, we design an enhanced feature pyramid network, utilizing deformable convolutions and feature fusion modules to enhance the feature information of rotating targets and improve the expressive capacity of features at all levels. The proposed algorithm in this paper performs well on the public DOTA dataset and HRSC2016 dataset, compared with other object detection methods, and the detection accuracy AP values of most object categories are improved by at least three percentage points. The results show that our method can accurately locate densely arranged and dynamically oriented targets, significantly reducing the risk of missing detections, and achieving higher levels of target detection accuracy.

A Remote Sensing Image Quality Interpretation Scale Characterization Method Based on the TTP Criterion

Accurate grading of remote sensing image interpretation is crucial for improving image classification and screening efficiency. Through extensive research, the General Image Quality Equation (GIQE) based on the National Imagery Interpretability Rating Scale (NIIRS) has been developed. However, poor reliability and low accuracy issues remain due to the failure to consider human visual characteristics. This paper introduces the Target Task Performance (TTP) criterion as a key parameter to reflect the cascading degradation factors of human visual perception characteristics and system imaging links, which improves the reliability of the model. A New Optimized Remote Sensing Image Quality Equation (NORSIQE), which effectively predicted the interpretability of image information, is constructed. Using 200 sets of test data, the quantitative relationship between key parameters (GSD, TTP, and SNR) of the NORSIQE and the subjective NIIRS level is obtained by a least squares regression fit, and the determination coefficient of the model is as high as 0.916. The model is evaluated for accuracy using 120 sets of validation data, showing an 87% improvement compared to the GIQE4. This method provides theoretical support for the development of new methods for remote sensing image quality evaluation and the design of payloads for remote sensing imaging systems.

U-Net-STN: A Novel End-to-End Lake Boundary Prediction Model

Detecting changes in land cover is a critical task in remote sensing image interpretation, with particular significance placed on accurately determining the boundaries of lakes. Lake boundaries are closely tied to land resources, and any alterations can have substantial implications for the surrounding environment and ecosystem. This paper introduces an innovative end-to-end model that combines U-Net and spatial transformation network (STN) to predict changes in lake boundaries and investigate the evolution of the Lake Urmia boundary. The proposed approach involves pre-processing annual panoramic remote sensing images of Lake Urmia, obtained from 1996 to 2014 through Google Earth Pro Version 7.3 software, using image segmentation and grayscale filling techniques. The results of the experiments demonstrate the model’s ability to accurately forecast the evolution of lake boundaries in remote sensing images. Additionally, the model exhibits a high degree of adaptability, effectively learning and adjusting to changing patterns over time. The study also evaluates the influence of varying time series lengths on prediction accuracy and reveals that longer time series provide a larger number of samples, resulting in more precise predictions. The maximum achieved accuracy reaches 89.3%. The findings and methodologies presented in this study offer valuable insights into the utilization of deep learning techniques for investigating and managing lake boundary changes, thereby contributing to the effective management and conservation of this significant ecosystem.

Formation and distribution of landslides controlled by thrust-strike-slip fault zones and fluvial erosion in the Western Qinling Mountains, China

In active orogenic belts, fault zones and fluvial erosion are the two principal factors that control the development of landslides. The Langma-Baiya fault zone (LBFZ) and the Bailongjiang fault zone (BFZ) in the Western Qinling Mountains (WQM) in China are part of an intracontinental orogenic zone, where landslides are the principal geomorphological process. It is also one of the regions with the most severe landslide disasters in China. We established a historical landslide database for the LBFZ and BFZ, based on high-precision remote sensing image interpretation and field investigation. The results indicated a total of 2764 landslides, including 1190 large landslides and 79 giant landslides. The quality of the landslide database was evaluated using the frequency-area curve distribution characteristics, and the results showed that the small-medium size landslides are underrepresented in the database, likely due to the influence of factors such as surface erosion, vegetation coverage, and human activities. Therefore, we focused on the spatial distribution and formation mechanism of large and giant landslides. Landslides are unevenly distributed in space; specifically, our field investigation and statistical analysis of swath profiles revealed that landslides are aggregated within tectonic zones, in areas of weak rock masses, and on the banks of rivers. Finally, based on the combination of the various factors controlling landslides, we divided the study area into three subregions, for which conceptual models were established. These models highlight the importance of the distance of the fault zone to a river, together with the degree of fluvial erosion, in controlling the development of landslides. The combined effects of fluvial erosion, geological structures, and the low-grade metamorphic rocks exposed along fault zones are major controls on the development of large landslides, and the local structural characteristics of thrust-strike-slip fault zones control the type of landslides.

Numerical Investigation of the Evolution Process of an Open-Pit Mine Landslide Using Discrete-Element Method

Because of the complexity of geological conditions and disturbance of external factors, the process of instability of open-pit landslides is very complicated. To illustrate this complex process, taking an open-pit coal mine landslide as an example, an accurate landslide geological section was established via detailed field investigation, remote sensing image interpretation, and field borehole data, in this study. A two-dimensional discrete-element model was established by the Fast CPU Matrix Computing of Discrete-Element Method (MatDEM, Version 1.60). The failure evolution of the landslide was investigated; then, the deformation behavior and dynamic characteristics of the landslide were systematically studied. The analyses of the failure process, velocity, displacement, and heat showed that the evolution of the landslide could be divided into the stages of initial deformation, accelerated deformation, and deceleration and stabilization. During the movement of the landslide, the gravitational potential energy of the landslide gradually converted into kinetic energy, heat, and elastic potential energy. The kinetic energy in the initial stage was small but rapidly increased and occupied the dominant position in the second stage. The kinetic energy gradually decreased and tended to be stable in the third stage. The final geometry and accumulation characteristics of the landslide obtained by numerical simulation were consistent with the corresponding field investigation and remote sensing interpretation. This work can provide a reference for the study of the evolution and prevention of open-pit mine landslides.