Archaeological site protection support system, archaeological site protection support method, and archaeological site protection support program

Abstract

This system provides support for archaeological site preservation activities, enabling the identification of undiscovered sites and buried structures, as well as the assessment of the impact of environmental changes and human activities on existing sites, which can then be used for the protection and restoration of those sites. [Solution] The system includes a point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning a potential archaeological site or an existing archaeological site; an image acquisition unit that acquires image data obtained by imaging a potential archaeological site or an existing archaeological site; a spectral image acquisition unit that acquires multispectral image data including infrared image data based on specific wavelengths reflected from the potential archaeological site or an existing archaeological site; a 3D position information acquisition unit that acquires position and orientation data, equipped with at least a GPS receiver and an inertial measuring instrument; and a mobile body on which these acquisition units are mounted. Based on the various data acquired while the mobile body is moving or while the mobile body is stationary, the system determines in real time whether an archaeological site exists or whether the impact of an existing archaeological site is assessed.

Description

Technical Field

[0005] , , ,

[0001] The present invention relates to a heritage protection activity support system, a heritage protection activity support method, and a heritage protection activity support program for supporting heritage protection activities including the discovery, excavation, protection, and restoration of heritage sites.

Background Art

[0002] In many cases, locations where there are existing heritage sites (heritage areas, structures) or potential heritage sites are physically difficult to access, and on-site surveys require a lot of time and cost. Also, conventionally, since records were left in the form of photos and drawings, it was difficult to capture three-dimensional details and fine textures, and it was impossible to completely reproduce the object. Furthermore, since existing heritage sites (heritage areas, structures) are at risk of changing or being lost due to natural disasters, environmental changes, and human activities, regular recording is essential. However, with conventional methods, this regular recording requires a great deal of time and labor and is difficult.

[0003] Therefore, conventionally, it has been proposed to scan the locus with high precision to create a digital three-dimensional model, and thereby save the accurate position and shape of the structure as digital data for use in analyzing the preservation state of the structure, managing the preservation environment, and taking deterioration countermeasures (see Patent Document 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, conventional methods have only analyzed archaeological sites and structures at the time a three-dimensional model was created, and considered management and countermeasures for these sites. This has not contributed to the discovery or excavation of further sites or structures, nor has it allowed for the identification of the impact of environmental changes or human activities on the sites. Therefore, from the perspective of protecting archaeological sites, conventional methods have the drawback of providing only limited information.

[0006] This invention has been made in view of the above circumstances, and its main objective is to provide a system, method, and program for supporting archaeological site protection activities that can be used for the protection and restoration of archaeological sites by identifying undiscovered archaeological sites and buried remains, and by confirming the impact of environmental changes and human activities on existing sites. [Means for solving the problem]

[0007] To achieve the above objectives, the archaeological site protection activity support system according to the present invention is a system that supports archaeological site protection activities, A point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning a potential archaeological site or an existing archaeological site, The image acquisition unit acquires image data obtained by imaging the aforementioned potential archaeological site or existing archaeological site, A spectral image acquisition unit that acquires multispectral image data including infrared image data based on specific wavelengths reflected from the aforementioned potential archaeological site or existing archaeological site, A 3D position information acquisition unit that acquires position and attitude data, comprising at least a GPS receiver and an inertial measuring instrument, A mobile body equipped with the point cloud data acquisition unit, the image acquisition unit, the spectral image acquisition unit, and the 3D position information acquisition unit, The acquired data storage unit stores the 3D point cloud data, image data, and multispectral image data of the potential archaeological site or existing archaeological site acquired while the mobile body is moving or stationary, as well as the position and orientation data at the time of data acquisition. Based on the data acquired by the aforementioned data acquisition unit, a detection and evaluation unit determines in real time whether there is a ruin or whether the impact of an existing ruin is assessed. It is characterized by possessing the following features.

[0008] Here, archaeological site protection activities include activities such as the discovery, excavation, protection, and restoration of the site. Furthermore, the mobile device may be a vehicle equipped with the above-mentioned acquisition components, an aerial vehicle including an unmanned aircraft such as a drone, or both. The appropriate choice of device should be made according to the environment of the access area, such as remote areas or mountainous regions. Furthermore, the identification of archaeological sites by the detection and evaluation unit includes the identification of undiscovered sites or buried structures, and the impact assessment of existing sites by the detection and evaluation unit includes the assessment of the impact of changes in the external environment on existing sites or the assessment of the impact of human activities on existing sites.

[0009] Therefore, based on 3D point cloud data, image data, and multispectral image data of potential or existing archaeological sites acquired while the mobile device is moving or stationary, as well as the position and orientation data of the mobile device at the time of data acquisition, the detection and evaluation unit determines in real time whether or not there are archaeological sites at potential sites, or assesses the impact of existing sites. This makes it possible to identify undiscovered sites and buried remains, and to take effective measures for the protection and restoration of archaeological sites while confirming the impact of environmental changes and human activities on existing sites (archaeological sites, remains, artifacts, etc.).

[0010] The determination of the presence of archaeological sites in the detection and evaluation unit may be performed by an archaeological site determination unit, which estimates the presence or absence of archaeological sites in a potential archaeological site by inputting input data, including 3D point cloud data acquired by the point cloud data acquisition unit and image data acquired by the image acquisition unit, into a learning model that has been pre-trained on the correlation between the presence or absence of archaeological sites and the input data. By using a learning model (AI) to determine the presence of archaeological sites in this way, it becomes possible to accurately identify sites that might be easily overlooked by visual inspection.

[0011] In this case, the input data for the learning model may further include multispectral image data acquired by the spectral image acquisition unit. By using multispectral image data, it becomes possible to detect features that cannot be identified with visible light, thereby improving the accuracy of determining the presence or absence of archaeological sites in potential archaeological sites, and also improving the accuracy of impact assessment of existing archaeological sites.

[0012] Furthermore, the detection and evaluation unit determines the impact assessment of existing ruins based on at least two of the 3D point cloud data acquired by the point cloud data acquisition unit, the image capture data acquired by the image capture unit, and the multispectral image data acquired by the spectral image acquisition unit, and constructs a digital twin that reproduces the existing ruins using digital twin technology, and A simulation execution unit that uses the aforementioned digital twin to simulate the impact of environmental changes and human activities on existing archaeological sites, It may be implemented in this manner. By using digital twins, it is possible to accurately simulate the impact of environmental changes and human activities on existing archaeological sites, making it possible to implement appropriate measures for the protection and restoration of these sites.

[0013] Furthermore, it is preferable to further include a visualization unit that visualizes the determination results from the detection and evaluation unit onto a 3D model formed using the 3D point cloud data, and a report generation unit that issues a report when the detection and evaluation unit determines that the ruins exist. This configuration allows the assessment results to be displayed in three dimensions, making them intuitively understandable and facilitating confirmation of discoveries and further investigation. Furthermore, since the existence of archaeological sites is reported, it becomes easier to confirm the presence of sites in potential archaeological locations. [Effects of the Invention]

[0014] As described above, according to the archaeological site protection activity support system, archaeological site protection activity support method, and archaeological site protection activity support program according to the present invention, while moving a moving body or while stopping the moving body, 3D point cloud data, captured image data, and multispectral image data of a potential archaeological site or an existing archaeological site obtained, and based on the position and orientation data of the moving body at the time of data acquisition, the detection evaluation unit determines in real time the presence of an archaeological site at the potential archaeological site or the impact evaluation of the existing archaeological site. Therefore, undiscovered archaeological sites and buried archaeological structures can be identified, and effective measures can be taken for the protection and restoration of archaeological sites while confirming the impact of environmental changes and human activities on existing archaeological sites.

Brief Description of the Drawings

[0015] [Figure 1] FIG. 8 shows an example in which the archaeological site protection activity support system according to the present invention is applied to a potential archaeological site and an archaeological site in a hilly area. [Figure 2] FIG. 11 is a block diagram showing a configuration example of the archaeological site protection activity support system according to the present invention. [Figure 3] FIG. 14 is a block diagram showing the configuration of a machine learning device used to determine the presence or absence of an archaeological site. [Figure 4] FIG. 17 is a flowchart showing an example of an operation process for determining the presence or absence of an archaeological site using a learning model. [Figure 5] FIG. 20 is a flowchart for explaining the construction steps of a system for evaluating the impact of an existing archaeological site. [Figure 6] FIG. 23 is a flowchart showing the flow of introducing and operating a system for evaluating the impact of an existing archaeological site.

Embodiments for Carrying Out the Invention

[0016] Hereinafter, embodiments according to the present invention will be described with reference to the accompanying drawings.

[0017] As an example in which the archaeological site protection activity support system S according to the present invention is applied, FIG. 1 shows an example of a potential archaeological site and an archaeological site with a raised-floor building site in a hilly area. As shown in Figure 2, the archaeological site protection support system S is equipped with a mobile unit M that has various sensors such as a LiDAR (Light Detection and Ranging) 1, an RGB camera 2, and a multispectral camera 3.

[0018] The mobile units M are envisioned to be manned or unmanned vehicles or mobile robots that move on the ground around the archaeological site, or flying vehicles such as drones or aircraft that fly over the archaeological site. Each mobile unit M is equipped with the aforementioned LiDAR1, RGB camera2, and multispectral camera3. Mapping using vehicles is known as MMS (Mobile Mapping System), and mapping using mobile robots is known as SLAM (Simultaneous Localization and Mapping). When these vehicles or mobile robots are used, the sensors mentioned above (LiDAR1, RGB camera2, multispectral camera3) are mounted on top of them. Data is then collected while these mobile devices M are moving or stationary (in the following example, a mobile robot R is used). Furthermore, aerial mapping systems using aircraft are known as Aerial Mapping Systems (UAV Mapping Systems), and they collect data using the aforementioned sensors mounted on aircraft such as drones and aircraft, either while moving or while hovering in a predetermined position (the following example shows the use of an unmanned aircraft, drone D).

[0019] Each of the above sensors is installed, for example, on the top of the main body of a mobile robot R, and for example, on the bottom of the main body of a drone D. These LiDAR1 and various cameras 2,3 should be able to rotate 360° around a vertical axis perpendicular to the support base surface of the mobile body M, and should also be installed so that their depression angle can rotate up and down within a predetermined range.

[0020] LiDAR1 works by shining a laser beam towards a potential or existing archaeological site, measuring the time it takes for the laser to reflect back, and using this data to determine the distance to the target object, thereby generating a 3D point cloud of the potential or existing archaeological site. In other words, it acquires 3D point cloud data generated by 3D scanning the potential or existing archaeological site using laser light. Based on this data, a detailed 3D model (3D map) of the potential or existing archaeological site is created.

[0021] The RGB camera (imaging device) 2 acquires image data generated by imaging potential or existing archaeological sites. It decomposes light reflected from the subject into three components—red, green, and blue—through an RGB filter. The camera's image sensor (CMOS or CCD) measures the intensity of each color of this decomposed light, converts it into a digital signal, and records it as the RGB value of each pixel. By combining the recorded RGB values, a full-color image is generated. This makes it possible to visually confirm physical deformations and color changes. For example, it becomes possible to visually confirm the surface shape of potential or existing archaeological sites, as well as fine details such as cracks, corrosion, and damage to existing sites.

[0022] The multispectral camera 3 can capture light in multiple wavelength bands, simultaneously acquiring information across a wide range of wavelengths, including not only visible light but also near-infrared and ultraviolet light. It also functions as an infrared sensor, acquiring multispectral image data (including infrared image data) based on specific wavelengths reflected from potential or existing archaeological sites. This camera makes it possible to identify invisible anomalies (anomalies that can be detected by temperature changes) such as overheating, cooling, and gas leaks at potential or existing archaeological sites.

[0023] Furthermore, the mobile unit M (mobile robot R or drone D) is equipped with a point cloud data acquisition unit 11 that acquires 3D point cloud data generated by 3D scanning a potential archaeological site or existing archaeological site using LiDAR 1, an image acquisition unit 12 that acquires image data generated by imaging a potential archaeological site or existing archaeological site using an RGB camera (imaging device) 2, and a spectral image acquisition unit 13 that acquires multispectral image data (including infrared image data) based on specific wavelengths reflected from the potential archaeological site or existing archaeological site using a multispectral camera 3 (including infrared camera 4). The 3D point cloud data, image data, and multispectral image data of the potential archaeological site or existing archaeological site are acquired while the mobile unit M is moving or while the mobile unit M is stationary.

[0024] Furthermore, the mobile body M is equipped with a 3D position information acquisition unit 15. This 3D position information acquisition unit 15 includes a GPS (Global Positioning System) unit and an inertial measurement unit (IMU). The GPS unit receives signals from satellites to acquire accurate position information of the mobile body. The IMU measures the movement of the mobile body M, such as acceleration, angular velocity, and direction of travel, using an accelerometer and a gyroscope.

[0025] The acquired data is then sent in real time via the data acquisition transmission unit 16 to the data center's management server 20 or cloud server via the communication network 10. Hereafter, the server to which the data is sent will be referred to as the management server 20. The management server 20 receives the transmitted acquired data with the acquired data receiving unit 21, and stores the acquired data received by the acquired data receiving unit 21 in the acquired data storage unit 22, associating it with the motion parameters and position information of the moving object M identified by the 3D position information acquisition unit 15. Then, based on the data stored in the acquired data storage unit 22, the detection and evaluation unit 23 determines whether or not there are archaeological sites (identifying undiscovered sites in potential archaeological sites, or identifying buried sites), and also determines the impact assessment of existing sites in real time based on the data stored in the acquired data storage unit 22 (assessing the impact of environmental changes on existing sites and assessing the impact of human activities on existing sites).

[0026] Identifying archaeological sites within this potential archaeological site and assessing the impact on existing sites (evaluating the impact of environmental changes on existing sites, and evaluating the impact of human activities on existing sites) can be done using various methods, one example of which will be explained below.

[0027] (Regarding the identification of archaeological sites: Regarding the identification of undiscovered or buried archaeological sites) First, I will explain how the detection and evaluation unit 23 identifies archaeological sites (determines whether or not archaeological sites exist) in potential archaeological sites. The following methods can be used to identify undiscovered or buried archaeological sites.

[0028] (Overview of methods for identifying archaeological sites) (1) Data collection The aforementioned data is collected using each of the sensors (LiDAR, RGB camera, GPS unit, and inertial measurement unit (IMU)). In other words, • LiDAR: Collects detailed 3D point cloud data of potential archaeological sites and existing archaeological sites. • RGB camera: Acquires high-resolution image data of potential and existing archaeological sites. • GPS unit: Provides accurate location information for mobile devices (such as mobile robots and drones). • IMU: Corrects the movement of moving objects and improves the accuracy of the data. (2) Data processing and integration • Data preprocessing system: Cleans and removes noise from collected raw data. This includes filtering algorithms and data correction techniques. • Data Integration System: Integrates LiDAR data and image data to generate high-precision 3D models. This utilizes image analysis techniques and data matching algorithms. (3) Building a learning model • Machine learning models: These are models used to learn the features of ruins and structures. Deep learning algorithms such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs) are used for this purpose. • Training dataset: The model is trained using data from existing archaeological sites and structures. The dataset includes 3D models and image data of known sites. • Feature extraction algorithms: Extract features of ruins and structures from collected 3D data. This includes techniques such as edge detection, shape recognition, and pattern matching. • Anomaly detection algorithm: Detects undiscovered archaeological sites and buried structures. Identifies areas that deviate from normal topographic and structural patterns, suggesting potential discoveries. The learning model used for AI analysis of data collected by the mobile device M, and the specific analysis methods, are described in detail below.

[0029] (Regarding the learning model) As shown in Figure 3, the machine learning device 30 includes an input data acquisition unit 31 that acquires data sets (information to be evaluated) obtained from 3D point cloud data generated by 3D scanning of the potential archaeological site with LiDAR1 and image data generated by imaging the potential archaeological site with RGB camera2 as input data; a label acquisition unit 32 that acquires data sets including the presence or absence of archaeological sites (which may also include the type and size of the archaeological sites) as labels; and a learning model construction unit 33 that constructs a learning model 35 by performing supervised learning using the input data and label pairs as training data. Here, the input data to be input to the input data acquisition unit 31 may be input directly from each of the sensors via the communication network 10, but it is preferable to use the data stored in the acquired data storage unit 22 of the management server 20 as input data.

[0030] Then, as shown in Figure 4, 3D point cloud data and image data of potential archaeological sites are acquired by various sensors mounted on the mobile unit M (step S41). Based on the acquired 3D point cloud data and image data of potential archaeological sites, the existence of archaeological sites is estimated using the constructed learning model 35 (corresponding to the archaeological site determination unit that estimates the presence or absence of archaeological sites: step S42). If the existence of archaeological sites is determined (step S43), this determination result (presence or absence of archaeological sites, type of archaeological site) is displayed visually (displayed on a 3D model or on a 2D map) or reported in report format (step S44).

[0031] (The learning model to be used) Here, suitable learning models to use include, for example, CNNs (Convolutional Neural Networks), RNNs (Recurrent Neural Networks), and hybrid neural networks including autoencoders, as described below.

[0032] (CNN: Convolutional Neural Networks) • Applications: Highly effective for feature extraction and pattern recognition from image data. It uses LiDAR data and high-resolution image data as input to analyze the shape and patterns of structures. • Structure: It consists of multiple convolutional layers, pooling layers, and fully connected layers. This extracts and classifies the spatial features of the data. The input and output data when using this learning model are as follows, for example. Input data: • LiDAR data: 3D point cloud data converted into 2D image formats (e.g., depth maps and elevation maps). It contains information such as the shape, depth, and height of the ground surface and structures. • High-resolution image data: Image data that shows the appearance of terrain and structures, such as aerial photographs and satellite images. Output data: • Extracted Feature Map: An intermediate output that extracts spatial features such as shape, edges, and patterns. This allows for a detailed understanding of the terrain and structures. • Classification results of structures: Based on the input data, the system classifies whether the structures correspond to known archaeological sites or natural landforms, or whether they are unknown artificial structures. Location information of anomalies: Identify the coordinates and area information of anomalies (potential locations of artificial objects) that deviate from the normal terrain pattern.

[0033] (RNN: Rent Neural Network) • Applications: Suitable for analyzing time-series data, it is used for analyzing continuous data collected by mobile mapping devices while they are in motion. For example, it can be used to analyze data continuously acquired from sensors and detect anomalies. • Configuration: Uses units such as LSTM (Long Short-Term Memory) and GRU (Gated Recurrent Unit) to capture long-term dependencies. The input and output data for this learning model are as follows, for example. Input data: • Time-series LiDAR data: Continuous data of 3D point cloud data collected by mobile mapping. Because it is acquired while moving, it includes continuously changing terrain information. Sequential image data: Aerial photographs and image data collected over time. This includes changes in viewpoint due to movement, as well as changes in structures and terrain. Output data: Anomaly detection results: Identifies deviations from normal terrain patterns and detects areas or points with abnormal changes. Predicting the existence of buried structures: Based on the time-series changes in anomaly points, we predict the location of structures that may be buried underground. We focus on areas where anomaly changes are continuous.

[0034] (Autoencoder) • Applications: Used for data dimensionality reduction and anomaly detection. By compressing high-dimensional LiDAR data into a low-dimensional latent space and reconstructing it, anomalies and characteristic patterns are extracted. • Configuration: It consists of two neural networks: an encoder and a decoder. The encoder compresses the data into a lower dimension, and the decoder reconstructs it. The input and output data when using this learning model are, for example, as follows: Input data: • High-dimensional LiDAR data: High-density point clouds and terrain data. This data is spatially complex, making dimensionality reduction effective. • High-resolution image data: Image data of structures and terrain. Used for feature compression and reconstruction for anomaly detection. Output data: • Low-dimensional latent representation: Feature quantities in the latent space obtained by compressing input data. They represent the essential patterns of the data and form the basis for anomaly detection. Reconstruction error: The difference between the input data and the reconstructed data. Areas with large reconstruction errors are detected as anomalies, and these anomaly areas are identified as showing patterns that deviate from the normal terrain. Location information of abnormal areas: Based on reconstruction errors, the coordinates of abnormal points or areas that deviate from the normal pattern are output.

[0035] (Specific analysis process) The 3D data analysis process using the above-described learning model is as follows: • Data preprocessing: Clean the collected 3D data and remove noise. Also, normalize and correct the data. • Feature extraction: Using CNNs, features are extracted from 3D and image data. Shapes, edges, patterns, etc., are detected. • Pattern matching: Compare with a database of known archaeological sites to identify matching patterns. Use an autoencoder to detect anomalies. • Anomaly detection: We also use RNNs to analyze continuous data and predict anomalies and potential hidden structures. • Integration and visualization of results: Integrate analysis results and visualize them using 3D visualization tools to make the results easy for users to understand. In this way, by using AI to analyze 3D data obtained from mobile mapping, it becomes possible to efficiently identify undiscovered ruins and structures.

[0036] Alternatively, undiscovered or buried archaeological sites can be identified using the following procedure. 1. Perform initial feature extraction and structure classification using a CNN. Initial Feature Extraction: First, features are extracted from terrain data and image data to detect the presence of artificial features and structures. Specifically, shapes, edges, and patterns are extracted from input 3D terrain data (such as LiDAR data) and high-resolution images using a CNN. This highlights anomalies, including artificial shapes and patterns. Structure Classification: Based on the extracted features, the structure is compared with a database of known archaeological sites to determine if it matches a known pattern. If the features differ from known archaeological sites or natural terrain, the site is marked as a candidate for an unknown artificial structure. Next, the CNN outputs the anomaly detected and candidate structural locations, which are then used for the next analysis.

[0037] 2. Time series data analysis and anomaly detection using RNNs We use an RNN to analyze the changes in time-series data regarding the locations of anomalies and potential structures marked by the CNN. Mobile mapping and continuously acquired data are used to track terrain information and the placement of structures that change over time, and to detect time-series anomalies that deviate from normal terrain patterns. If a particular anomaly appears repeatedly, it suggests the possibility of structures buried beneath the surface. The RNN results, which identify points with a high probability of being continuous anomalies or buried structures over time, are then passed on to the next autoencoder analysis.

[0038] 3. Dimensionality Reduction and Anomaly Detection using Autoencoders To compress high-dimensional terrain data into a lower-dimensional space and identify anomalies using reconstruction errors, an autoencoder is then used to compress the high-dimensional data related to the anomalies extracted by CNNs or RNNs (e.g., detailed point cloud information and fine terrain features) into a lower-dimensional latent space. By comparing the input data with the reconstructed data and detecting areas with large reconstruction errors, features that differ from the normal pattern are highlighted. Areas with large reconstruction errors may be artificial structures and are therefore identified as anomalies or distinctive topographic features. The autoencoder outputs location information for abnormal areas with large reconstruction errors, and this is integrated and visualized as the final identification result. By applying the learning models described above and linking the outputs from each model to subsequent analyses, the accuracy of detecting undiscovered archaeological sites and buried structures is improved. Finally, by integrating the outputs from the three models, more reliable candidate sites for investigation are identified. 4. Integration and visualization of results (1) Integration of results The outputs of the CNN, RNN, and autoencoder (anomalies, candidate structures, consecutive anomalies, and anomalies due to reconstruction errors) are integrated to ultimately identify potential locations of undiscovered ruins or buried structures. (2) Visualization and reporting of results • 3D Visualization Tool: The integrated results are visualized using a 3D visualization tool in the visualization processing unit 24 and presented in an easy-to-understand format for the user. This allows users to confirm the location and details of discovered ruins and structures, intuitively identify areas requiring excavation, and improve the efficiency of the survey. • Interactive map: The visualization processing unit 24 displays discovered ruins and structures on a map, providing a user-navigable interface. • Report generation system: The report generation unit 25 generates a detailed report of the analysis results. This report includes an overview of the discovery, location information, and screenshots of the 3D model.

[0039] As described above, by using a learning model, it becomes possible to identify undiscovered trajectories and buried ruins, enabling real-time analysis: when a mobile mapping device collects data, it can simultaneously perform AI analysis and provide results immediately (real-time analysis). Furthermore, by having the AI ​​model learn the patterns of normal terrain and structures and accurately identify deviations from them, it becomes possible to efficiently find undiscovered archaeological sites (high-precision anomaly detection). Furthermore, the entire process, from data collection to analysis and reporting of results, is automated, enabling it to handle large-scale surveys (automation and scalability). The systems described above can not only accelerate the discovery of unknown ruins and structures, but can also be used to conduct detailed investigations of known ruins.

[0040] To identify undiscovered ruins and structures, a method combining mapping by a mobile device M with the above-mentioned learning model offers the following advantages over conventional methods. (1) A dramatic improvement in data volume and data accuracy Traditional methods primarily involved manual topographic surveys and aerial photograph analysis. These methods have limited data collection ranges, and their accuracy varies depending on location and conditions. This method utilizes a mobile object mapping technique to rapidly collect high-precision, wide-area 3D data. Data obtained using LiDAR sensors and high-resolution cameras provides detailed information that cannot be obtained with conventional methods. (2) Advanced analytical capabilities of AI Traditional method: Human experts analyze data visually and manually. This is time-consuming and prone to human error. This method uses AI to analyze large amounts of data in a short time, enabling the detection of subtle features and patterns that are often overlooked by the human eye. Deep learning models can automatically learn and identify the features of complex terrains and structures. (3) Real-time analysis and feedback Traditional method: Data collection and analysis are separated, and data is brought back from the field for analysis, resulting in a time lag. Immediate response at the site is difficult. This method combines a mobile mapping device with an AI analysis system, enabling real-time data collection and analysis. It allows for immediate identification of anomalies and potential archaeological sites on-site, facilitating rapid response. (4) Automation and scalability Traditional methods: Large-scale surveys required significant human resources and time. Surveys covering vast areas were particularly inefficient. This method automates the entire process from data collection to analysis and reporting, enabling efficient surveys even over large geographical areas. AI-powered analysis is scalable and can process large amounts of data simultaneously. (5) Advanced anomaly detection and pattern recognition Traditional methods: Because they rely on human judgment, it is difficult to identify subtle anomalies or buried structures. Inferences based on known ruins and structures are the mainstream approach. This method: The AI ​​model is trained on a large dataset and excels at anomaly detection and pattern recognition. It can identify unknown patterns and subtle anomalies with high accuracy. (6) Advanced visualization and user interface Conventional methods: Visualization of results was limited, and 2D maps and reports were the norm. This method uses a 3D visualization tool to display results in three dimensions, allowing users to understand them intuitively. This makes it easier to confirm findings and conduct further investigations. As described above, this method, which combines AI and mobile mapping technology, is characterized by high-precision data collection and analysis, automated processes, real-time feedback, and advanced anomaly detection capabilities, making it possible to efficiently and accurately identify undiscovered ruins and buried structures.

[0041] In the above, data from LiDAR and a high-resolution RGB camera were used as input data to form the learning model, but multispectral image data captured by a multispectral camera may also be added as further input data. By collecting multispectral data, including infrared and ultraviolet light in addition to visible light, it may be possible to detect different features of the Earth's surface and subsurface. In this way, by introducing a multimodal learning model that integrates and analyzes LiDAR data, image data, and multispectral data, it becomes possible to perform analysis using richer information.

[0042] Alternatively, generative models such as GANs (Generative Adversarial Networks) can be used to generate new shapes and patterns of ruins from existing data, thereby predicting unknown ruins. Furthermore, self-supervised learning, which learns features from unlabeled data, can be introduced to efficiently utilize large amounts of unlabeled data.

[0043] Furthermore, while the above configuration involved real-time identification of archaeological sites via a management server over a network, edge computing capabilities could be incorporated into the data collection device itself (mobile unit) to facilitate real-time analysis at the site. This would enable more instantaneous data analysis and feedback. Furthermore, in the example above, interactive 3D visualization may be used for the analysis results (prediction results), providing an interface that allows users to intuitively manipulate the 3D data. For example, a zoom function to examine the discovered structures in detail and a function to view from different viewpoints may be added. Furthermore, virtual reality (VR) and augmented reality (AR) technologies could be used to allow users to explore the ruins in a virtual space.

[0044] (Regarding the assessment of the impact on existing archaeological sites (assessment of the impact of environmental changes on existing archaeological sites, assessment of the impact of human activities on existing archaeological sites)) Next, we will describe in detail a method for predicting the impact of environmental changes and human activities on existing archaeological sites based on data collected by the sensors mounted on the aforementioned mobile device M.

[0045] 1. System Overview (Data collection and digital twin generation) • Real-time data acquisition: High-resolution LiDAR sensors and multispectral cameras are used to collect data in real time. By using drones and ground robots in combination, high-precision data can be obtained even in hard-to-reach locations. • Data integration and advanced 3D modeling: Integrate different types of data to generate a detailed digital twin of the site. This allows for an accurate understanding of the current state of the site. (Environmental monitoring and predictive simulation) • Real-time monitoring of environmental changes: Weather sensors and environmental sensors are used to monitor temperature, humidity, precipitation, soil moisture, pH levels, air quality, etc., in real time. • Combined Simulation: By simulating a combination of environmental changes and human impacts, we can predict the impact on archaeological sites with high accuracy. (AI analysis and anomaly detection) • Anomaly detection and early warning: Using AI, anomalies are detected from simulation results, and an early warning system is built. If an anomaly is detected, an alert is immediately issued, and countermeasures are proposed. • Generating optimal restoration scenarios: Using AI models, the system proposes the most suitable restoration methods for deterioration and damage. This enables efficient protection and restoration of archaeological sites. (User interface and visualization) • Interactive 3D visualization: Provides an interface that allows users to manipulate the digital twin and intuitively understand the simulation results. Implements zoom functionality and the ability to view from different viewpoints. • Introduction of VR / AR technology: Virtual reality (VR) and augmented reality (AR) will be used to visually review virtual tours of the ruins and restoration scenarios. This will make it easier for engineers and conservation specialists to develop concrete restoration plans.

[0046] 2. System Construction Steps The steps for constructing a system for assessing the impact of existing archaeological sites are as follows (see Figure 5). (Phase 1: Survey and Design: S51) • Requirements definition: Define the specific functional requirements of the system and identify the necessary technologies and resources. • Prototyping: Design a prototype of each module and perform proof-of-concept. (Phase 2: Infrastructure Construction: S52) • Construction of a data acquisition system: Integrate high-resolution LiDAR sensors, multispectral cameras, drones, and ground robots to build a data acquisition system. • Development of a digital twin generation module: Develop software to integrate collected data and generate a high-precision digital twin. (Phase 3: Environmental Monitoring and Predictive Simulation: S53) • Installation of environmental monitoring sensors: Install weather sensors and environmental sensors to collect data in real time. • Development of simulation algorithms: Develop and test algorithms for simulating environmental changes and human impacts. (Phase 4: AI analysis and anomaly detection: S54) • Training anomaly detection algorithms: Based on the collected data, train anomaly detection algorithms and integrate them into the system. • Development of AI for generating repair scenarios: Develop an AI model that proposes the optimal repair method for deterioration and damage. (Phase 5: User Interface and Visualization: S55) • Development of interactive 3D visualization tools: Develop tools that allow users to manipulate digital twins and intuitively understand simulation results. • Development of VR / AR applications: Develop VR / AR applications that allow users to visually check virtual tours and restoration scenarios. (Phase 6: Continuous Improvement and Evaluation: S56) • Regular evaluation and feedback: Conduct regular evaluations of the system, collect user feedback, and implement improvements. • Technology updates: Continuously consider introducing new technologies and methods to improve system performance.

[0047] 3. Specific System Configuration (1) Data collection Of the aforementioned sensors, the following data is collected using LiDAR1 and the multispectral camera3. • LiDAR: Acquires 3D point cloud data to be used to generate detailed 3D models of existing archaeological sites. • Multispectral camera: Acquires multispectral image data to be used to analyze the surface condition and material properties of archaeological sites. (2) Generation of a digital twin • Data Integration and Processing: Integrate different types of collected data (3D point cloud data and multispectral image data) to generate a high-precision digital twin. This allows for the detailed reconstruction of the structure and materials of the archaeological site. A digital twin can be created using any two of the following: 3D point cloud data acquired by LiDAR1, imaging data acquired by RGB camera2, or multispectral image data acquired by multispectral camera3. Here, however, we use 3D point cloud data and multispectral image data, which are effective for analyzing the properties and state of materials. (3) Obtain weather and environmental information • Weather information: Weather data such as temperature, humidity, and precipitation are collected in real time by weather sensor 17. • Environmental information: Soil moisture, pH value, and ambient air quality are obtained using environmental sensor 18. (4) Conducting predictive simulations Hybrid Simulation Model: This model implements a hybrid simulation combining physics-based simulations with data-driven AI models. This enables highly accurate and efficient simulations. • Predictive maintenance: Using anomaly detection algorithms, deterioration and damage to archaeological sites are detected early, enabling predictive maintenance. This allows for the long-term protection of the sites. • Scenario-based simulation: Simulate multiple scenarios (different environmental conditions, human impacts, and remediation methods) to propose the optimal protection and remediation strategy. Unlike conventional archaeological site preservation and restoration systems, this system, with its unique features and configuration, offers new value by leveraging advanced digital twin technology and predictive simulations. This system enables more precise and effective decision-making regarding the preservation and restoration of archaeological sites. (Regarding environmental change simulations) Environmental change simulations simulate the impact of weather changes (temperature, humidity, wind speed, etc.) on existing archaeological sites. Specifically, they use weather data measured by weather sensors (temperature, humidity, precipitation, etc.) and environmental data measured by environmental sensors (soil moisture, pH value, surrounding air quality, etc.) to predict the impact on existing archaeological sites. The details are as follows: • Data collection The weather sensor 17 collects data such as temperature, humidity, and wind speed. Environmental sensors 18 collect data such as soil moisture, pH value, and air quality. • Construction of a physical model Create a physical model that takes into account the material and structural characteristics of the ruins. We set up equations to simulate physical phenomena such as heat conduction, moisture absorption and release, and erosion. • Run the simulation Based on the collected environmental data, a physical model will be used to simulate the impact on the archaeological site. We set up scenarios for short-term weather variability (e.g., seasonal temperature changes) and long-term climate change (e.g., global warming), and then run simulations.

[0048] (Regarding simulations of human-induced impacts) This study simulates the impact of human activities such as the influx of tourists and construction work on existing historical sites. Specifically, it will cover the following: • Data collection A human flow sensor 19, such as cameras, will be installed to monitor tourist influx data and pedestrian movement patterns. Collect information regarding construction activities and other related matters in the surrounding area. • Model building Create a pedestrian flow model that takes into account the movement patterns and length of stay of tourists. Create a model to simulate physical effects such as vibration and gravity. • Run the simulation Using the above model, we will simulate the load on specific areas of the ruins based on tourist influx data. Furthermore, the impact of vibrations and ground subsidence caused by construction activities will be simulated to predict the effects on the archaeological site.

[0049] (Structural deterioration simulation) Structural deterioration simulation: This predicts how the structure of an archaeological site will deteriorate over time. Specifically, it is as follows: • Data collection We will collect current data on the archaeological site using high-resolution LiDAR1 and multispectral cameras3. We will also collect data on past deterioration and restoration history. • Training of AI models To learn degradation patterns, a machine learning model is trained using historical data. We will construct regression models and generative models (GANs) to predict the rate of degradation and its impact. • Run the simulation Using current data and AI models, we predict which parts of the ruins will deteriorate and how over time. We set up different environmental conditions and protective measures scenarios, and perform degradation predictions for each scenario. (5) Analysis and visualization of simulation results (Anomaly detection) The simulation results are analyzed to detect abnormal degradation or impacts. If an anomaly is detected, an alert is issued, and a detailed analysis is performed. (Proposed restoration scenario) Based on the simulation results, we propose the optimal repair method. This allows for efficient execution of specific repair procedures and material selection. (visualization) • Interactive 3D visualization The simulation results are visualized as an interactive 3D model. Users can manipulate the digital twin of the ruins to intuitively understand the progression of deterioration (simulation results) and restoration scenarios. • Utilization of VR / AR technology VR Tours: These tours offer a virtual tour of the ruins, allowing users to visually confirm restoration scenarios. Users wear a VR headset and experience the simulation results in real time, visually confirming the progression of deterioration and restoration scenarios. This makes it easier for engineers and conservation specialists to develop concrete restoration plans. In this case, an AR application could be used, allowing users to scan the ruins on-site using smartphones or tablets and overlay the simulation results, thereby enabling rapid decision-making on-site.

[0050] 4. Examples of specific simulation scenarios (Scenario 1: Prediction of deterioration due to seasonal environmental changes) • Winter (cold season) simulation This simulation will examine the effects of temperature drops and freeze-thaw cycles on the stonework and structure of archaeological sites. It will predict how cracking and delamination of the stonework will progress due to expansion caused by freezing and contraction caused by thawing. • Summer (high temperature period) simulation This study simulates the effects of high temperatures and dryness on the materials of archaeological sites. It predicts expansion and contraction due to temperature changes, as well as weakening due to drying. • Rainy season (high humidity period) simulation This study simulates the effects of high humidity and rainfall on the structure of archaeological sites. It predicts the risk of deterioration, corrosion, and mold growth due to moisture penetration. (Scenario 2: Prediction of deterioration due to influx of tourists) • Tourism period simulation This study simulates the impact of large crowds during the tourist season on archaeological sites. It predicts the load on specific areas by considering people's movement patterns and dwell times. This allows for the evaluation of deterioration due to compaction, wear, and vibration. • Event simulation This simulation will show the impact of special events and festivals on archaeological sites. It will predict the effects of temporary high loads associated with these events on the structure of the sites, allowing for proactive countermeasures to be taken. (Scenario 3: Predicting the impact of construction activities) • Simulation of the impact of nearby construction projects This project simulates the effects of nearby construction activities on archaeological sites, including vibrations and ground subsidence. It also predicts the physical impact of heavy machinery use and excavation work on the sites. This allows for adjustments to construction plans and consideration of protective measures. (Scenario 4: Predicting the effectiveness of the repair method) • Material selection simulation Simulate the effects of different repair materials (e.g., specific mortars or waterproofing agents). Evaluate the durability and environmental compatibility of each material to assist in selecting the optimal material. • Repair method simulation The effects of different repair methods (e.g., injection reinforcement, surface protection) are simulated. The effectiveness and cost of each method are compared, and the optimal repair plan is proposed.

[0051] 5. Technical Details and Implementation (Modeling and Simulation) • Finite Element Method (FEM): The finite element method is used to simulate structural degradation and physical effects with high accuracy. • Computational Fluid Dynamics (CFD): CFD is used to simulate the effects of weather conditions and humidity changes on the archaeological site. (AI and machine learning) • Anomaly detection algorithm: An anomaly detection algorithm (e.g., an autoencoder for anomaly detection) is used to detect abnormal degradation patterns. • Predictive Model: A predictive model using time series data (e.g., LSTM, GRU) is used to predict future degradation progression. • Generative Models (GANs): Generative Adversarial Networks (GANs) are used to simulate the progression of degradation and the state after repair. This allows for the generation of more realistic scenarios. • Regression models: Use regression models (e.g., linear regression, random forest regression) to predict the rate and impact of degradation. (Data processing and integration) • Data preprocessing: Cleans up collected sensor data and historical degradation data, and converts it into a format suitable for simulation. • Data Integration Platform: Build a data platform to integrate different data sources and perform advanced data analysis and simulations. For example, using Apache Hadoop or Spark. (Real-time monitoring and feedback) • Real-time data streaming: Collect data in real time from weather and environmental sensors and feed it back into simulation models. Use Apache Kafka or AWS Kinesis. • Dashboards: Build dashboards to display simulation results and real-time monitoring data. Use Grafana and Tableau to achieve intuitive data visualization.

[0052] 6. System Implementation Example (Data collection and integration) • LiDAR1: Conducts a 3D scan of the entire site to generate a high-precision digital twin. LiDAR data is stored in point cloud format. • Environmental sensor 18: Collects weather data (temperature, humidity, wind speed, precipitation) and soil data. This data is streamed to a database in real time. (Modeling and Simulation) • Finite Element Method (FEM) Model: The structure of the archaeological site is subdivided, and physical properties are defined for each part. FEM software (e.g., ANSYS, COMSOL) is used to simulate the effects of environmental changes and loads. • AI model training: Anomaly detection algorithms and regression models are trained using historical, degraded data. Python libraries (e.g., TensorFlow, PyTorch) are used. (Running the simulation) • Environmental change simulation: Using collected weather data and FEM models, simulate the effects of seasonal variations and climate change. Different weather conditions are set for each scenario to evaluate the impact on the archaeological site. • Human-induced impact simulation: Tourist movement data is input into a pedestrian flow model to simulate the load on specific areas. Special loads during events are also taken into consideration. • Structural deterioration simulation: Using current data and AI prediction models, this simulation will show how the archaeological site deteriorates over time. Different environmental conditions and restoration method scenarios will be set up, and deterioration predictions will be performed for each. (Analysis and visualization of results) • Anomaly detection and alerts: Analyze simulation results to detect abnormal degradation or impacts. If an anomaly is detected, an alert will be displayed on the dashboard. • Proposal of repair scenarios: Based on the simulation results, the AI ​​model proposes the optimal repair method. It specifically outlines the repair procedure and material selection. • Interactive 3D visualization: Simulation results are displayed as a 3D model. Users can interact with the digital twin to visually understand the progression of deterioration and repair scenarios. • VR / AR applications: Virtual tours using VR headsets and on-site verification using AR applications on smartphones are possible. This supports rapid decision-making on-site.

[0053] The process for implementing and operating this system is shown in Figure 6. (Initial data collection: S61) • Sensor deployment: High-resolution LiDAR sensors and environmental sensors are deployed around the ruins by moving the mobile unit M, and data collection begins. • Baseline data acquisition: Collect current data on the site to obtain foundational data for generating a digital twin. (Modeling and simulation preparation: S62) • Creating a digital twin: Based on the collected data, a highly accurate 3D model of the ruins will be created. • Construction of physical models: Create physical models that take into account the material and structural characteristics of the ruins, and import them into FEM or CFD software. (Simulation run: S63) • Environmental change simulation: Based on collected environmental data, we simulate the effects of seasonal variations and long-term climate change. • Human impact simulation: Using data on tourist influx and construction activities, we simulated the impact on the archaeological site. • Structural deterioration simulation: Using current data and AI models, we simulate how the ruins deteriorate over time. (Analysis and visualization of results: S64) • Anomaly detection and alerts: Analyzes simulation results to detect abnormal degradation or impacts. If an anomaly is detected, an alert will be displayed on the dashboard. • Proposal of repair scenarios: Based on the simulation results, the AI ​​model proposes the optimal repair method. It specifically outlines the repair procedure and material selection. • Interactive 3D visualization: Simulation results are displayed as a 3D model. Users can interact with the digital twin to visually understand the progression of deterioration and repair scenarios. • VR / AR applications: Virtual tours using VR headsets and on-site verification using AR applications on smartphones are possible. (Continuous improvement and evaluation: S65) • Regular evaluation and feedback: Conduct regular evaluations of the system, collect user feedback, and implement improvements. • Technology updates: Continuously consider introducing new technologies and methods to improve system performance.

[0054] (Examples of application to archaeological sites) Next, we will describe examples of its application to archaeological sites. (1) Application example 1: Protection of ancient temples • Data acquisition: A detailed 3D model of the temple was created using a high-resolution LiDAR1 and a multispectral camera3. • Environmental simulation: Simulate seasonal variations in the area where the temple is located and evaluate the effects of cold and hot periods. • Human-induced impact simulation: Collect tourist movement data and simulate the load on specific areas. The impact during tourist seasons and events is also evaluated. • Deterioration Simulation: Based on past deterioration data and current environmental data, the simulation shows how the stone and wood materials of the temple will deteriorate. • Proposal of restoration scenarios: Based on simulation results, we propose the optimal restoration method and materials. We create a report including restoration procedures and cost evaluations. • Visualization and Feedback: A 3D visualization of the temple's digital twin will be created to allow administrators and conservation professionals to understand it intuitively. On-site verification will also be conducted using VR / AR applications. (2) Application example 2: Protection of ancient city walls • Data collection: Drones and LiDAR sensors were used to create a complete 3D model of the castle walls. Environmental sensors were installed to collect weather data. • Environmental simulation: Simulate the effects of seasonal variations and climate change, and identify areas where weathering and erosion are particularly advanced. • Structural deterioration simulation: Predict how the city walls will deteriorate based on past repair history and current deterioration data. Simulate different repair scenarios and evaluate the optimal approach. • Human impact simulation: Simulate the impact of nearby construction activities and the influx of tourists on the city walls. Assess the risks of vibration and ground subsidence. • Restoration scenario proposal: Based on simulation results, the optimal restoration method and materials are proposed. Restoration priorities and cost evaluations are also provided. • Visualization and Feedback: Simulation results are visualized in 3D using a digital twin of the city walls. Results are provided in a format that is easy for administrators and conservation professionals to understand, and on-site verification is also conducted using VR / AR applications. (3) Application example 3: Protection of ancient cemeteries • Data acquisition: A detailed 3D model of the cemetery was created using a high-resolution LiDAR1 and a multispectral camera3. Environmental sensors18 were installed to monitor humidity and pH levels. • Environmental simulation: Simulates the effects of seasonal temperature fluctuations and precipitation, and particularly evaluates the impact of water infiltration and soil changes on gravestones. • Structural deterioration simulation: Based on the material of the tombstone and past deterioration data, the simulation shows how it deteriorates over time. The effectiveness of different protective measures is also evaluated. • Human-induced impact simulation: Simulating the impact of tourist influx and local activities on the cemetery. Particularly evaluating wear and tear caused by walking and contact. • Restoration scenario proposal: Based on simulation results, the optimal restoration method and materials are proposed. Restoration priorities and cost evaluations are also provided. • Visualization and Feedback: Simulation results are visualized in 3D using a digital twin of the cemetery. Results are provided in a format that is easy for administrators and conservation professionals to understand, and on-site verification is also conducted using VR / AR applications.

[0055] (summary) Leveraging digital twins and predictive simulation technologies enables a new approach to the protection and restoration of archaeological sites. By utilizing specific simulation methods to accurately predict environmental changes and human impacts, optimal protection and restoration strategies can be formulated. Furthermore, the use of interactive 3D visualization and VR / AR technologies allows all stakeholders to share simulation results in an intuitively understandable format, supporting effective decision-making. This makes long-term protection and sustainable management of cultural heritage possible.

[0056] Furthermore, the aforementioned archaeological site protection activity support system S can also be provided in the form of a program (archaeological site protection activity support program) that causes a computer to execute each step of the archaeological site protection activity support method described above. [Explanation of Symbols]

[0057] 11. Point cloud data acquisition unit 12 Image acquisition unit 13. Spectral image acquisition unit 31 Input data acquisition unit 32 Label acquisition unit 33 Learning Model Construction Department 35 Learning Models S Archaeological Site Protection Support System M Mobile object R-type mobile robot F flying object

Claims

1. A system to support archaeological site preservation activities, A point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning a potential archaeological site or an existing archaeological site, The image acquisition unit acquires image data obtained by imaging the aforementioned potential archaeological site or existing archaeological site, A spectral image acquisition unit that acquires multispectral image data including infrared image data based on specific wavelengths reflected from the aforementioned potential archaeological site or existing archaeological site, A 3D position information acquisition unit that acquires position and attitude data, comprising at least a GPS receiver and an inertial measuring instrument, A mobile body equipped with the point cloud data acquisition unit, the image acquisition unit, the spectral image acquisition unit, and the three-dimensional position information acquisition unit, The acquired data storage unit stores the 3D point cloud data, image data, and multispectral image data of the potential archaeological site or existing archaeological site acquired while the mobile body is moving or stationary, as well as the position and orientation data at the time of data acquisition. Based on the data acquired by the aforementioned data acquisition unit, a detection and evaluation unit determines in real time whether there is a ruin or whether the impact of an existing ruin is assessed. A system for supporting archaeological site preservation activities, characterized by being equipped with the following features.

2. The archaeological site protection support system according to claim 1, characterized in that the mobile body includes at least one of a mobile robot or an aerial body, the identification of the archaeological site by the detection and evaluation unit includes the identification of an undiscovered archaeological site or the identification of a buried structure, and the impact assessment of the existing archaeological site by the detection and evaluation unit includes the assessment of the impact of changes in the external environment on the existing archaeological site or the assessment of the impact of human activities on the existing archaeological site.

3. The determination of the existence of the archaeological site by the detection and evaluation unit is as follows: The archaeological site protection support system according to claim 1, characterized in that input data including 3D point cloud data acquired by the point cloud data acquisition unit and image capture data acquired by the image capture unit is input into a learning model that has been pre-trained to determine the correlation between the presence or absence of archaeological sites and the input data, thereby inferring the presence or absence of archaeological sites in the potential archaeological site.

4. The aforementioned input data further includes multispectral image data acquired by the spectral image acquisition unit, according to claim 3, for the archaeological site protection activity support system.

5. The determination of the impact assessment of existing archaeological sites by the aforementioned detection and evaluation unit is as follows: A digital twin construction unit constructs a digital twin of the existing ruins using digital twin technology, based on at least two of the 3D point cloud data acquired by the point cloud data acquisition unit, the image capture data acquired by the image capture unit, and the multispectral image data acquired by the spectral image acquisition unit. A simulation execution unit that uses the aforementioned digital twin to simulate the impact of environmental changes and human activities on existing archaeological sites, The archaeological site protection activity support system according to claim 1, characterized in that it is implemented by [the specified method].

6. A visualization unit visualizes the determination result from the detection and evaluation unit onto a 3D model formed using the 3D point cloud data, When the presence of the archaeological site is determined by the detection and evaluation unit, a report issuance unit issues a report, The archaeological site protection activity support system according to claim 1, further comprising the features described above.

7. A method of supporting archaeological site preservation activities, A data acquisition step involves storing data acquired by each acquisition unit while a mobile body equipped with the following components is in motion or stationary: a point cloud data acquisition unit that acquires 3D point cloud data obtained by 3D scanning a potential archaeological site or an existing archaeological site; an image acquisition unit that acquires image data obtained by imaging the potential archaeological site or an existing archaeological site; a spectral image acquisition unit that acquires multispectral image data including infrared image data based on specific wavelengths reflected from the potential archaeological site or an existing archaeological site; and a 3D position information acquisition unit that acquires position and attitude data, equipped with at least a GPS receiver and an inertial measuring instrument; Based on the acquired data, a detection and evaluation step is performed to determine in real time whether there is a ruin or an impact assessment of an existing ruin, A method for supporting archaeological site preservation activities, characterized by comprising the following:

8. The determination of the existence of the archaeological site in the aforementioned detection and evaluation step is, The input data, including 3D point cloud data acquired by the point cloud data acquisition unit and image capture data acquired by the image capture unit, is input into a learning model that has been pre-trained to determine the correlation between the presence or absence of archaeological sites and the input data. This is performed in an archaeological site determination step that estimates the presence or absence of archaeological sites in the potential archaeological site based on the input data. The method for supporting archaeological site protection activities as described in item 7.

9. The method for supporting archaeological site protection activities according to claim 8, further comprising multispectral image data acquired by the spectral image acquisition unit as the input data.

10. The determination of the impact assessment of existing archaeological sites in the aforementioned detection and evaluation step is as follows: A digital twin construction step involves constructing a digital twin that recreates the existing ruins using digital twin technology, based on at least two of the 3D point cloud data acquired by the point cloud data acquisition unit, the image capture data acquired by the image capture unit, and the multispectral image data acquired by the spectral image acquisition unit. A simulation execution step in which the impact of environmental changes and human activities on existing archaeological sites is simulated using the aforementioned digital twin, The method for supporting archaeological site protection activities according to claim 7, characterized in that it is carried out by [the specified method].

11. A visualization step in which the determination result from the detection and evaluation step is visualized on a 3D model formed using the 3D point cloud data, If the existence of the archaeological site is determined by the detection and evaluation step, a report issuance step is performed to issue a report, The method for supporting archaeological site protection activities according to claim 7, further comprising the features described above. method.

12. A program for supporting archaeological site protection activities that causes a computer to perform each step of the method for supporting archaeological site protection activities described in any one of claims 7 to 11.

Images