CN117934737A - Intelligent generation method for ancient cultural relic digital map - Google Patents

Abstract

The present invention discloses an intelligent generation method of digital maps of ancient cultural relics, including the following method steps: S1, using a multimodal sensor on an unmanned aerial vehicle platform to perform remote sensing photography or laser scanning on ancient cultural relics to obtain image data or point cloud data; S2, using deep learning and computer vision technology to filter, align, segment, match and identify image data or point cloud data; S3, using augmented reality and virtual reality technology to perform real and dynamic virtual restoration and display of feature points or surface models; S4, using cloud computing and big data technology to store, analyze and share ancient cultural relics data. The present invention can improve the efficiency, accuracy, clarity and realism of digital mapping of ancient cultural relics, and by using artificial intelligence technologies such as deep learning and computer vision, automatically and intelligently process image data or point cloud data to extract high-quality and high-density feature points or surface models.

Description

Intelligent Generation Method of Digital Map of Ancient Cultural Relics

Technical Field

The present invention relates to the technical field of map generation, and in particular to a method for intelligently generating a digital map of ancient cultural relics.

Background technique

Digital archaeology has become a major trend in the development of archaeology in the information age. This technology takes the acquisition and mapping of spatial information and cultural relics information at the archaeological site as its starting point, and combines it with spatial analysis and simulation of historical and cultural heritage, virtual reality display, database construction, etc. to achieve digital preservation, transmission and sharing, thereby providing information support for archaeological research, protection of cultural relics and historical sites, and archiving of various data.

The amount of data in the existing technology is large. Since immovable cultural relics resources are often widely distributed, numerous, and complex in form, the amount of data collected is also very large, which brings huge challenges to data storage, transmission, and processing. The data accuracy in the existing technology is low. Since the collected data has problems such as noise, occlusion, and missing, the feature points or surface models obtained after processing may have errors or distortions, and cannot completely restore the true state of the cultural relics. Data dissemination in the existing technology is difficult. Since immersive technology requires specific equipment and environment, such as head-mounted displays, handles, trackers, etc., there are certain restrictions on user usage conditions, which is not conducive to the widespread dissemination and popularization of cultural relic data. Therefore, how to provide an intelligent generation method for digital maps of ancient cultural relics is a problem that technicians in this field urgently need to solve.

Summary of the invention

One purpose of the present invention is to propose an intelligent generation method for digital maps of ancient cultural relics. The present invention can improve the efficiency, accuracy, clarity and realism of digital mapping of ancient cultural relics. By utilizing artificial intelligence technologies such as deep learning and computer vision, image data or point cloud data can be automatically and intelligently processed, such as filtering, alignment, segmentation, matching, recognition, etc., to extract high-quality, high-density feature points or surface models.

According to an embodiment of the present invention, a method for intelligently generating a digital map of ancient cultural relics includes the following steps:

S1. Use drone platforms equipped with multimodal sensors to perform remote sensing photography or laser scanning of ancient artifacts to obtain image data or point cloud data;

S2. Use deep learning and computer vision technology to filter, register, segment, match and identify image data or point cloud data;

S3. Use augmented reality and virtual reality technologies to perform real and dynamic virtual restoration and display of feature points or surface models;

S4. Use cloud computing and big data technologies to store, analyze and share ancient cultural relics data.

Optionally, S1 specifically includes performing control measurement, establishing a control network and control points, performing remote sensing photography or laser scanning, obtaining image data or point cloud data of cultural relics, performing image processing or point cloud processing, extracting feature points or surface models, performing data conversion and output, and generating a digital map or a three-dimensional model.

Optionally, the control measurement includes establishing a spatial resection equation or an ortho transformation equation according to known control point coordinates and corresponding image point coordinates, and solving unknown parameters or coordinates by the least squares method:

Among them, (x, y, z) are the spatial coordinates of the object point, (X, Y, Z) are the spatial coordinates of the photography center, f is the focal length of the camera, (a ₁ , b ₁ , c ₁ ) are the image space coordinates of the image point, r _ij is the element of the rotation matrix, (x ^′ , y ^′ ) is the image plane coordinate of the image point, (x, y) is the ground plane coordinate of the object point, a _i and b _i are the ortho transformation parameters;

The remote sensing photography or laser scanning includes using a remote sensing camera or a laser scanner to image or scan the cultural relics to obtain image data or point cloud data of the cultural relics:

I(x,y)＝f(L(x,y),R(x,y),G(x,y),B(x,y));

P(x,y,z)＝g(L(x,y,z),R(x,y,z),G(x,y,z),B(x,y,z));

Among them, I(x,y) is the image data, f is the image function, L(x,y) is the brightness value, R(x,y), G(x,y), B(x,y) are the red, green and blue color values, P(x,y,z) is the point cloud data, and g is the point cloud function;

The image processing or point cloud processing includes filtering, registering, segmenting, matching, and extracting feature points or surface models by image processing or computer vision techniques.

I ^′ (x, y) = h(I(x, y), k);

P ^′ (x, y, z) = i(P(x, y, z), k);

Among them, I ^′ (x, y) is the processed image data, h is the image processing function, k is the parameters such as filter or transformation matrix, P ^′ (x, y, z) is the processed point cloud data, and i is the computer vision function;

The data conversion and output includes converting the feature points or surface models using interpolation, fitting, and projection techniques to generate a digital map or a three-dimensional model:

Z(x,y)＝j(X(x,y),Y(x,y),Z(x,y));

M(x,y,z)＝k(X(x,y,z),Y(x,y,z),Z(x,y,z));

Among them, Z(x,y) is the digital map, j is the interpolation or fitting function, M(x,y,z) is the three-dimensional model, and k is the projection or transformation function.

Optionally, the filtering includes denoising, smoothing, and enhancing the image data or point cloud data, and the registration includes aligning the image data or point cloud data at different viewing angles or at different times to the same coordinate system:

Among them, R and t are rotation matrices and translation vectors, representing the transformation parameters from source data to target data, and p _i and q _i are corresponding points in the source data and target data;

The segmentation includes dividing the image data or point cloud data into regions or objects with similar properties or semantic meanings:

Among them, S is the segmentation result, which means that the data is divided into k subsets, _Si represents the i-th subset, and _Ci represents the center or representative point of the i-th subset.

The matching involves finding similar or repeated patterns or structures in the image data or point cloud data:

Where I is the image data, T is the template, and (x, y) is the optimal position of the template in the image data;

The identification includes classifying or labeling areas or objects in the image data or point cloud data:

y＝f(x；w);

Among them, y is the category of the data, x is the feature of the data, w is the parameter of the classifier, and f is the function of the classifier.

Optionally, S3 includes using computer graphics, human-computer interaction, and winter media methods to perform realistic and dynamic virtual restoration and display of feature points or surface models, simulating the appearance and behavior of ancient artifacts, and allowing users to interact with them.

Optionally, the virtual restoration includes reconstructing the three-dimensional geometric shape of the ancient artifact using feature points or surface models, and restoring the texture, color, and lighting details of the ancient artifact based on historical data or expert opinions:

M = f(I,P);

M ^′ = g(M,T);

Where M is the 3D model, f is the restoration function based on the image, I is the image data, P is the point cloud data, M ^′ is the restored 3D model, g is the restoration function based on the model, and T is the template or model;

The virtual display includes rendering, animation, and sound processing of the restored three-dimensional model using computer graphics, human-computer interaction, and media methods, and presenting it to users through different devices or platforms:

I ^′ (x, y) = h(I(x, y), M, R, t);

I″(x,y)＝i(M,R,t);

Among them, I ^′ (x, y) is the display result based on augmented reality, h is the display function of augmented reality, I(x, y) is the image data of the real environment, M is the three-dimensional model, R and t are the rotation matrix and translation vector, representing the transformation parameters from the three-dimensional model to the real environment, I″(x, y) is the display result based on virtual reality, and i is the display function of virtual reality.

Optionally, the S4 includes utilizing the Internet, virtualization, focus groups, and parallel computing methods to integrate computing resources and data resources dispersed in different locations to form a distributed system to provide users with various services and applications.

Optionally, the storage includes safe and stable preservation and backup of the ancient artifact data:

B＝{b ₁ ,b ₂ ,…,b _n }；

F＝{f ₁ ,f ₂ ,…,f _n }；

O＝{o ₁ ,o ₂ ,…,o _n }；

Among them, B is a block-based storage set, _bi is the i-th block, F is a file-based storage set, _fi is the i-th file, O is an object-based storage set, o _i is the i-th object.

Optionally, the analysis includes processing and mining ancient artifact data:

y = f(x);

y＝g(x,s);

y＝h(x,w);

Among them, y is the analysis result, x is the input data, f is the analysis function based on batch processing, g is the analysis function based on stream processing, s is the state or valley parameter, h is the analysis function based on machine learning, and w is the model or weight parameter.

Optionally, the sharing includes exchanging and transmitting the ancient artifact data:

Q＝{q ₁ ,q ₂ ,…,q _n }；

T＝{t ₁ ,t ₂ ,…,t _n }；

S = {s ₁ ,s ₂ ,…,s _n };

Among them, Q is a shared set based on message queues, _qi is the i-th queue, T is a shared set based on publish-subscribe, _ti is the i-th topic or channel, S is a shared set based on service bus, _si is the i-th middleware or proxy;

The visualization includes graphical and interactive display of ancient artifact data:

G = {g ₁ ,g ₂ ,…, _gn };

M＝{m ₁ ,m ₂ ,…,m _n };

N＝{n ₁ ,n ₂ ,…, _nn };

Among them, G is a graph-based visualization set, _gi is the i-th graph, M is a map-based visualization set, _mi is the i-th map, N is a network-based visualization set, _ni is the i-th network.

The beneficial effects of the present invention are:

(1) The present invention can improve the efficiency, accuracy, clarity and realism of digital mapping of ancient cultural relics. By using an unmanned aerial vehicle platform equipped with a multimodal sensor, all-round, multi-angle, and multi-band remote sensing photography or laser scanning of ancient cultural relics can be performed to obtain high-resolution, high-definition, and high-accuracy image data or point cloud data. By using artificial intelligence technologies such as deep learning and computer vision, the image data or point cloud data can be automatically and intelligently processed, such as filtering, registration, segmentation, matching, and recognition, to extract high-quality and high-density feature points or surface models.

(2) The present invention can realize multi-dimensional, multi-level, and multi-form display and interaction of ancient cultural relic data. By utilizing immersive technologies such as augmented reality and virtual reality, the feature points or surface models can be realistically and dynamically restored and displayed virtually, simulating the appearance and behavior of ancient cultural relics, allowing users to immersively experience the world of ancient cultural relics and interact with them. By utilizing distributed technologies such as cloud computing and big data, ancient cultural relic data can be securely and stably stored, analyzed, and shared, realizing functions such as integration, query, statistics, and visualization of ancient cultural relic data, and supporting access from different platforms and devices.

(3) The present invention can promote the scientificity and standardization of the protection and restoration of ancient cultural relics. By establishing a digital archive of ancient cultural relics at the mineral level, it can provide comprehensive, systematic and scientific records and information for ancient cultural relics, and provide a reliable basis and reference for the protection and restoration of ancient cultural relics. By realizing the digital transformation and reuse of ancient cultural relic data, it can provide new means and channels for the protection and restoration of ancient cultural relics.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

FIG1 is a flow chart of a method for intelligently generating a digital map of ancient cultural relics proposed by the present invention;

FIG. 2 is a flow chart of an ancient cultural relic digital map intelligent generation method of an ancient cultural relic digital map proposed by the present invention.

Detailed ways

The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

Referring to Figures 1-2, a method for intelligently generating a digital map of ancient cultural relics includes the following steps:

S1. Use drone platforms equipped with multi-modal sensors to conduct all-round, multi-angle, and multi-band remote sensing photography or laser scanning of ancient cultural relics to obtain high-resolution, high-definition, and high-accuracy image data or point cloud data;

In this embodiment, S1 specifically includes performing control measurements, establishing control networks and control points, performing remote sensing photography or laser scanning, obtaining image data or point cloud data of cultural relics, performing image processing or point cloud processing, extracting feature points or surface models, performing data conversion and output, and generating digital maps or three-dimensional models.

In this embodiment, the control measurement includes establishing a spatial resection equation or an ortho transformation equation based on known control point coordinates and corresponding image point coordinates, and solving unknown parameters or coordinates by the least squares method:

Among them, (x, y, z) are the spatial coordinates of the object point, (X, Y, Z) are the spatial coordinates of the photography center, f is the focal length of the camera, (a ₁ , b ₁ , c ₁ ) are the image space coordinates of the image point, r _ij is the element of the rotation matrix, (x ^′ , y ^′ ) is the image plane coordinate of the image point, (x, y) is the ground plane coordinate of the object point, a _i and b _i are the ortho transformation parameters;

Remote sensing photography or laser scanning involves using remote sensing cameras or laser scanners to image or scan cultural relics and obtain image data or point cloud data of cultural relics:

I(x,y)＝f(L(x,y),R(x,y),G(x,y),B(x,y));

P(x,y,z)＝g(L(x,y,z),R(x,y,z),G(x,y,z),B(x,y,z));

Among them, I(x,y) is the image data, f is the image function, L(x,y) is the brightness value, R(x,y), G(x,y), B(x,y) are the red, green and blue color values, P(x,y,z) is the point cloud data, and g is the point cloud function;

Image processing or point cloud processing includes filtering, registering, segmenting, matching, and extracting feature points or surface models by using image processing or computer vision techniques:

I ^′ (x, y) = h(I(x, y), k);

P ^′ (x, y, z) = i(P(x, y, z), k);

Among them, I ^′ (x, y) is the processed image data, h is the image processing function, k is the parameters such as filter or transformation matrix, P ^′ (x, y, z) is the processed point cloud data, and i is the computer vision function;

Data conversion and output include using interpolation, fitting, and projection techniques to convert feature points or surface models to generate digital maps or three-dimensional models:

Z(x,y)＝j(X(x,y),Y(x,y),Z(x,y));

M(x,y,z)＝k(X(x,y,z),Y(x,y,z),Z(x,y,z));

Among them, Z(x,y) is the digital map, j is the interpolation or fitting function, M(x,y,z) is the three-dimensional model, and k is the projection or transformation function.

S2. Use artificial intelligence technologies such as deep learning and computer vision to automatically and intelligently process image data or point cloud data, such as filtering, registration, segmentation, matching, and recognition, to extract high-quality, high-density feature points or surface models;

In this embodiment, filtering includes denoising, smoothing, and enhancing operations on image data or point cloud data. Filtering: Filtering refers to denoising, smoothing, and enhancing operations on image data or point cloud data to improve data quality and visualization effects.

Registration involves aligning image data or point cloud data from different perspectives or at different times to the same coordinate system to achieve data fusion and comparison. Registration can be divided into two categories: feature-based registration and density-based registration. Feature-based registration refers to the use of feature points or feature descriptors to find the correspondence between data, such as SIFT, SURF, etc. Density-based registration refers to the use of iterative closest point algorithm or its variants to minimize the distance error between data:

Among them, R and t are rotation matrices and translation vectors, representing the transformation parameters from source data to target data, and p _i and q _i are corresponding points in the source data and target data;

Segmentation involves dividing image data or point cloud data into regions or objects with similar properties or semantic meanings:

Among them, S is the segmentation result, which means that the data is divided into k subsets, _Si represents the i-th subset, and _Ci represents the center or representative point of the i-th subset.

Matching involves finding similar or repeated patterns or structures in image data or point cloud data. Matching can be divided into two categories: feature-based matching and template-based matching. Feature-based matching refers to using feature points or feature descriptors to measure the similarity between data, such as RANSAC, Hough transform, etc. Template-based matching refers to using predefined templates or models to search for target locations in data, such as correlation matching, shape matching, etc.:

Where I is the image data, T is the template, and (x, y) is the optimal position of the template in the image data;

Recognition includes classifying or labeling areas or objects in image data or point cloud data to achieve semantic understanding of the data. Recognition is divided into two categories: rule-based recognition and learning-based recognition. Rule-based recognition refers to the use of artificially defined rules or logic to determine the category of data, such as decision trees, expert systems, etc. Learning-based recognition refers to the use of machine learning or deep learning methods to learn data features and classifiers from a large amount of training data, such as support vector machines and convolutional neural networks:

y＝f(x；w);

Among them, y is the category of the data, x is the feature of the data, w is the parameter of the classifier, and f is the function of the classifier.

S3. Using immersive technologies such as augmented reality and virtual reality to realistically and dynamically restore and display feature points or surface models, simulate the appearance and behavior of ancient artifacts, and allow users to immersively experience the world of ancient artifacts and interact with them;

In this embodiment, S3 includes immersive technologies such as augmented reality and virtual reality, which refers to the use of computer graphics, human-computer interaction, winter media and other methods to perform real and dynamic virtual restoration and display of feature points or surface models, simulate the appearance and behavior of ancient artifacts, and allow users to immersively experience the world of ancient artifacts and interact with them.

In this implementation, virtual restoration refers to the reconstruction of the three-dimensional geometric shape of ancient artifacts using feature points or surface models, and the restoration of the texture, color, lighting and other details of ancient artifacts based on historical data or expert opinions to achieve a sense of reality and history. Image-based restoration refers to the use of image data or point cloud data as input to generate a three-dimensional model through image processing or computer vision methods, such as multi-view stereo reconstruction, structured light reconstruction, etc. Model-based restoration refers to the use of predefined templates or models as input to adapt to three-dimensional data through deformation, fusion, optimization and other methods, such as free deformation, mesh deformation, etc.:

M = f(I,P);

M ^′ = g(M,T);

Where M is the 3D model, f is the restoration function based on the image, I is the image data, P is the point cloud data, M ^′ is the restored 3D model, g is the restoration function based on the model, and T is the template or model;

Virtual display includes virtual display, which refers to the use of computer graphics, human-computer interaction, and media to render, animate, and sound effects the restored three-dimensional model, and present it to users through different devices or platforms to achieve immersion and interaction. Display based on augmented reality refers to superimposing the three-dimensional model into the real environment and dynamically adjusting it according to the user's perspective or position, such as plane tracking, space tracking, etc. Display based on virtual reality refers to placing the user in a virtual environment composed entirely of three-dimensional models, and performing real-time response according to the user's head or hand movements, such as head tracking, gesture recognition, etc.:

I ^′ (x, y) = h(I(x, y), M, R, t);

I″(x,y)＝i(M,R,t);

Among them, I ^′ (x, y) is the display result based on augmented reality, h is the display function of augmented reality, I(x, y) is the image data of the real environment, M is the three-dimensional model, R and t are the rotation matrix and translation vector, representing the transformation parameters from the three-dimensional model to the real environment, I″(x, y) is the display result based on virtual reality, and i is the display function of virtual reality.

S4. Use distributed technologies such as cloud computing and big data to safely and stably store, analyze and share ancient cultural relic data, realize functions such as integration, query, statistics, and visualization of ancient cultural relic data, and support access from different platforms and devices.

In this implementation, S4 includes distributed technologies such as computing and big data, which refers to the use of the Internet, virtualization, focus groups, parallel computing and other methods to integrate computing sources and data sources scattered in different locations to form a flexible, scalable and efficient distributed system to provide users with various services and applications.

In this embodiment, storage includes safe and stable preservation and backup of ancient artifact data to prevent data loss or damage:

B＝{b ₁ ,b ₂ ,…,b _n }；

F＝{f ₁ ,f ₂ ,…,f _n }；

O = {o ₁ ,o ₂ ,…,o _n };

Among them, B is a block-based storage set, _bi is the i-th block, F is a file-based storage set, _fi is the i-th file, O is an object-based storage set, o _i is the i-th object.

In this implementation, analysis involves processing and mining ancient artifact data to extract valuable information and knowledge:

y = f(x);

y＝g(x,s);

y＝h(x,w);

Among them, y is the analysis result, x is the input data, f is the analysis function based on batch processing, g is the analysis function based on stream processing, s is the state or valley parameter, h is the analysis function based on machine learning, and w is the model or weight parameter.

In this implementation, sharing includes exchanging and transmitting ancient cultural relic data to achieve the integration and sharing of data resources:

Q＝{q ₁ ,q ₂ ,…,q _n }；

T＝{t ₁ ,t ₂ ,…,t _n }；

S = {s ₁ ,s ₂ ,…,s _n };

Among them, Q is a shared set based on message queues, _qi is the i-th queue, T is a shared set based on publish-subscribe, _ti is the i-th topic or channel, S is a shared set based on service bus, _si is the i-th middleware or proxy;

Visualization includes graphical and interactive display of ancient artifact data to improve the comprehensibility and operability of the data:

G = {g ₁ ,g ₂ ,…, _gn };

M＝{m ₁ ,m ₂ ,…,m _n };

N＝{n ₁ ,n ₂ ,…, _nn };

Among them, G is a graph-based visualization set, _gi is the i-th graph, M is a map-based visualization set, _mi is the i-th map, N is a network-based visualization set, _ni is the i-th network.

Digital maps of ancient cultural relics refer to the use of digital surveying and mapping technology to digitally map immovable cultural relics resources, obtain the spatial location and distribution information of cultural relics, and convert them into digital maps to facilitate users to browse and locate at different scales.

Based on the intelligent generation method of digital maps of ancient cultural relics provided in the document, I can conceive a specific implementation example for you to simulate a complete digital map generation process of ancient cultural relics and provide some data to prove its beneficial effects.

Embodiment 1:

Step 1: Data Collection

A site was selected as the object. A drone equipped with a high-resolution multimodal sensor was used to conduct multiple flights at different time periods to conduct all-round, multi-angle remote sensing photography and laser scanning of the ancient city site. Detailed image data and point cloud data were obtained, including the topography, structure and details of the site.

Step 2: Data processing

The collected data is processed using deep learning and computer vision technology, including filtering to remove noise, registration to correct the alignment between data, segmentation to identify different structural elements, matching and recognition to extract feature points. High-quality, high-density feature points and surface models are formed, laying the foundation for virtual restoration.

Step 3: Virtual restoration and display

Using augmented reality and virtual reality technologies, the feature points and surface models are realistically and dynamically restored to simulate the historical appearance and behavior of the ancient city ruins. Users can immersively experience the historical environment and structure of the ancient city through virtual reality helmets or augmented reality applications.

Step 4: Data storage and sharing

Cloud computing and big data technologies are used to safely store, analyze and share the generated data. The data of ancient cultural relics can be integrated, queried, counted and visualized, and can be accessed from multiple platforms and devices.

Improved efficiency: Traditional surveying of ancient artifacts may take weeks, but using this method, it only takes a few days from data collection to virtual restoration.

Accuracy and clarity: Through high-precision sensors and advanced image processing technology, we achieve feature point accuracy of up to 0.5 cm and image clarity reaching HD levels.

Realism and user experience: In the VR environment, user feedback shows that they have a deeper understanding and experience of the historical environment of the ancient city.

Through this embodiment, we can not only effectively preserve and display the information of ancient cultural relics and sites, but also provide an immersive historical experience, which is conducive to the protection and popularization of ancient cultural relics education.

The present invention can improve the efficiency, accuracy, clarity and realism of digital mapping of ancient cultural relics. By utilizing an unmanned aerial vehicle platform equipped with a multimodal sensor, all-round, multi-angle, and multi-band remote sensing photography or laser scanning of ancient cultural relics can be performed to obtain high-resolution, high-definition, and high-accuracy image data or point cloud data. By utilizing artificial intelligence technologies such as deep learning and computer vision, image data or point cloud data can be automatically and intelligently processed, such as filtering, registration, segmentation, matching, and recognition, to extract high-quality, high-density feature points or surface models.

The present invention can realize multi-dimensional, multi-level, and multi-form display and interaction of ancient cultural relic data. By utilizing immersive technologies such as augmented reality and virtual reality, feature points or surface models can be realistically and dynamically virtually restored and displayed, and the appearance and behavior of ancient cultural relics can be simulated, allowing users to immersively experience the world of ancient cultural relics and interact with them. By utilizing distributed technologies such as cloud computing and big data, ancient cultural relic data can be safely and stably stored, analyzed, and shared, realizing functions such as integration, query, statistics, and visualization of ancient cultural relic data, and supporting access from different platforms and devices.

The present invention can promote the scientificity and standardization of the protection and restoration of ancient cultural relics. By establishing a digital archive of ancient cultural relics at the mineral level, it can provide comprehensive, systematic and scientific records and information for ancient cultural relics, and provide a reliable basis and reference for the protection and restoration of ancient cultural relics. By realizing the digital transformation and reuse of ancient cultural relic data, it can provide new means and channels for the protection and restoration of ancient cultural relics.

The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims (10)

1. The intelligent generation method of the ancient relic digital map is characterized by comprising the following steps of:

S1, carrying a multi-mode sensor by using an unmanned plane platform, and carrying out remote sensing photography or laser scanning on ancient cultural relics to obtain image data or point cloud data;

S2, filtering, registering, dividing, matching and identifying the image data or the point cloud data by utilizing deep learning and computer vision technology;

S3, utilizing the augmented reality and virtual reality technologies to conduct real and dynamic virtual restoration and display on the feature points or the surface model;

And S4, storing, analyzing and sharing the ancient cultural relic data by utilizing cloud computing and big data technology.

2. The intelligent generation method of the digital map of the ancient cultural relics according to claim 1, wherein the step S1 specifically comprises the steps of performing control measurement, establishing a control network and control points, performing remote sensing photography or laser scanning, obtaining image data or point cloud data of the cultural relics, performing image processing or point cloud processing, extracting characteristic points or surface models, performing data conversion and output, and generating a digital map or a three-dimensional model.

3. The method of claim 2, wherein the controlling the measurement includes creating a space back intersection equation or an orthographic transformation equation according to the known coordinates of the control point and the coordinates of the corresponding image point, and solving the unknown parameters or coordinates by a least square method:

Wherein, (X, Y, Z) is the spatial coordinates of the object point, (X, Y, Z) is the spatial coordinates of the photographing center, f is the focal length of the camera, (a ₁,b₁,c₁) is the image space coordinates of the image point, rij is an element of the rotation matrix, (X ', Y') is the image plane coordinates of the image point, (X, Y) is the ground plane coordinates of the object point, a _i and b _i are the orthographic transformation parameters;

The remote sensing photography or laser scanning comprises the steps of imaging or scanning the cultural relics by using a remote sensing camera or a laser scanner to obtain image data or point cloud data of the cultural relics:

I(x,y)＝f(L(x,y),R(x,y),G(x,y),B(x,y))；

P(x,y,z)=g(L(x,y,z)，R(x,y,z)，G(x,y,z)，B(x,y,z))；

wherein I (x, y) is image data, f is an image function, L (x, y) is a luminance value, R (x, y), G (x, y), B (x, y) is a red, green, blue tristimulus value, P (x, y, z) is point cloud data, and G is a point cloud function;

The image processing or point cloud processing comprises the steps of filtering, registering, segmenting and matching image data or point data by utilizing an image processing or computer alkaline vision technical means, and extracting characteristic points or surface models:

I′(x,y)＝h(I(x,y),k)；

P′(x,y,z)＝i(P(x,y,z),k)；

Wherein I '(x, y) is processed image data, h is an image processing function, k is parameters such as a filter or a transformation matrix, P' (x, y, z) is processed point cloud data, and I is a computer vision function;

The data conversion and output comprises the steps of carrying out data conversion on the characteristic points or the surface model by utilizing interpolation, fitting and projection technical means to generate a digital map or a three-dimensional model:

Z(x，y)＝j(X(x，y),Y(x,y),Z(x,y))；

M(x,y,z)＝k(X(x,y,z)，Y(x，y，z)，Z(x,y,z))；

Where Z (x, y) is a digital map, j is an interpolation or fitting function, M (x, y, Z) is a three-dimensional model, and k is a projection or transformation function.

4. The method for intelligently generating the digital map of the ancient cultural relics according to claim 1, wherein the filtering comprises denoising, smoothing and enhancing the image data or the point cloud data, and the registering comprises registering the image data or the point cloud data with different angles of view or different times under the same coordinate system:

Wherein R and t are rotation matrices and translation vectors representing transformation parameters from source data to target data, and p _i and q _i are corresponding points in the source data and the target data;

The segmentation includes dividing the image data or point cloud data into regions or objects having similar properties or semantic meanings:

Where S is the segmentation result, representing the division of the data into k subsets, S _i represents the ith subset, c _i represents the center or representative point of the ith subset

The matching includes finding similar or repetitive patterns or structures in the image data or point cloud data:

wherein I is image data, T is a template, and (x, y) is the optimal position of the template in the image data;

the identifying includes classifying or labeling regions or objects in the image data or point cloud data:

y＝f(x；w)；

where y is the class of data, x is the feature of the data, w is the parameter of the classifier, and f is the function of the classifier.

5. The intelligent generation method of the digital map of the ancient relic according to claim 1, wherein the step S3 comprises the steps of virtually and dynamically restoring and displaying the feature points or the surface model by using computer graphics, man-machine interaction and winter media method, and simulating the appearance and the behavior of the ancient relic to enable a user to interact with the ancient relic.

6. The method for intelligent generation of digital map of ancient relics according to claim 6, wherein the virtual restoration comprises reconstructing three-dimensional geometry of the ancient relics using feature points or surface models, and restoring texture, color, illumination details of the ancient relics according to historic data or expert opinion:

M＝f(I,P)；

M′＝g(M,T)；

Wherein M is a three-dimensional model, f is a restoration function based on an image, I is image data, P is point cloud data, M' is a restored three-dimensional model, g is a restoration function based on a model, and T is a template or model;

The virtual exhibition comprises rendering, animation and sound effect processing of the restored three-dimensional model by utilizing computer graphics, man-machine interaction and a named media method, and presenting the three-dimensional model to a user through different devices or platforms:

I′(x,y)＝h(I(x,y),M,R,t)；

I″(x,y)＝i(M，R，t)；

Wherein I' (x, y) is a display result based on augmented reality, h is a display function based on augmented reality, I (x, y) is image data of a real environment, M is a three-dimensional model, R and t are rotation matrices and translation vectors representing transformation parameters from the three-dimensional model to the real environment, I "(x, y) is a display result based on virtual reality, and I is a display function based on virtual reality.

7. The intelligent generation method of the ancient document digital map according to claim 1, wherein S4 comprises integrating the computing theft sources and the data theft sources scattered at different places by using internet, virtualization, jiao Qun and parallel computing methods to form a distributed system, and providing various services and applications for users.

8. The method for intelligent generation of an ancient document digital map according to claim 7, wherein said storing comprises safe and stable storage and backup of the ancient document data:

B＝{b₁,b₂,...,b_n}；

F＝{f₁,f₂,...,f_n}；

O＝{o₁,o₂,...,o_n}；

where B is the block-based storage set, B _i is the ith block, F is the file-based storage set, F _i is the ith file, O is the object-based storage set, and O _i is the ith object.

9. The method for intelligent generation of an ancient document digital map according to claim 7, wherein the analyzing comprises processing and mining of the ancient document data:

y＝f(x)；

y＝g(x，s)；

y＝h(x，w)；

Where y is the analysis result, x is the input data, f is the analysis function based on batch processing, g is the analysis function based on stream processing, s is the state or valley parameters, h is the analysis function based on machine learning, and w is the model or weight parameters.

10. The method for intelligent generation of an ancient document digital map according to claim 7, wherein the sharing comprises exchanging and transmitting the ancient document data:

Q＝{q₁,q₂,...,q_n}；

T＝{t₁,t₂,...,t_n}；

S＝{s₁,s₂,...,s_n}；

Wherein Q is a shared set based on a message queue, Q _i is an ith queue, T is a shared set based on a publish-subscribe, T _i is an ith topic or channel, S is a shared set based on a service bus, and S _i is an ith middleware or agent;

The visualization includes a graphical and interactive presentation of the historic data:

G＝{g₁,g₂,...,g_n}；

M＝{m₁,m₂,...,m_n}；

N＝{n₁,n₂,...,n_n}；

Where G is the graph-based visualization set, G _i is the ith graph, M is the map-based visualization set, M _i is the ith map, N is the network-based visualization set, and N _i is the ith network.