Method and device for visual identification of Chinese ancient buildings and storage medium

By collecting images of the external outline, internal load-bearing structure, and external painted texture features of traditional Chinese buildings, and using differential hashing and LBP algorithms for preprocessing, combined with a multi-channel convolutional neural network, the problem of not being able to accurately distinguish the age and specifications of traditional Chinese buildings in existing technologies has been solved, achieving high-accuracy recognition.

CN116486249BActive Publication Date: 2026-11-10BEIJING XINGHAN GUYUN TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202210022250.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-01-10
Publication Date
2026-11-10
Estimated Expiration
2042-01-10

AI Technical Summary

Technical Problem

Existing technology cannot accurately distinguish the age and specifications of traditional Chinese buildings. It mainly relies on the geographical location and shape of the building and cannot effectively utilize the features of the internal load-bearing components and the external painted textures.

Method used

Images of the external outline, internal load-bearing structure, and external painted texture of traditional Chinese buildings were collected. These images were preprocessed using differential hashing and LBP algorithms, and then combined with a multi-channel convolutional neural network for feature combination and recognition.

Benefits of technology

It enables precise differentiation of the age and specifications of traditional Chinese buildings, improving the accuracy of identification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116486249B_ABST
    Figure CN116486249B_ABST
Patent Text Reader

Abstract

The application relates to the technical field of image intelligent identification, and discloses a method for visual identification of Chinese ancient buildings, which comprises the following steps: collecting characteristic images of the whole building and multiple local parts of the building, and labeling the image types of the characteristic images; pre-processing the characteristic images by using corresponding methods according to the image types to obtain pre-processed images; combining the pre-processed images into characteristic combination data, and labeling the characteristic combination data with the characteristic identification types to which the characteristic combination data belong; and analyzing the characteristic combination data belonging to the same characteristic identification type by using corresponding convolutional neural networks according to the characteristic identification types, so as to obtain the identified building. The method can accurately distinguish the ages and specifications of Chinese ancient buildings according to the external contour, internal load-bearing structure and external color drawing texture of the Chinese ancient buildings. The application further discloses a device for visual identification of Chinese ancient buildings and a storage medium.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of image intelligent recognition technology, specifically to a method, apparatus and storage medium for visual recognition of traditional Chinese architecture. Background Technology

[0002] Currently, building identification primarily involves collecting data on a building's geographical location and external features, then using neural networks to identify the building. However, traditional Chinese buildings, besides their external shape, exhibit distinct characteristics in their internal load-bearing components, structural form, and external painted decorations, reflecting their specific era. Therefore, when buildings have similar external shapes, it is necessary to differentiate their age and specifications based on their internal load-bearing components and external painted textures.

[0003] Using neural networks to identify buildings based on their geographical location and shape is insufficient to accurately distinguish the age and specifications of traditional Chinese buildings. Summary of the Invention

[0004] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0005] This disclosure provides a method, apparatus, and storage medium for visual recognition of traditional Chinese buildings. Based on the external outline, internal load-bearing structure, and external painted texture of traditional Chinese buildings, the age and specifications of traditional Chinese buildings can be accurately distinguished.

[0006] In some embodiments, the method includes: acquiring feature images of the entire building and multiple parts of the building, and labeling the feature images with image types; preprocessing the feature images according to the image types using appropriate methods to obtain preprocessed images; combining the preprocessed images into feature combination data and labeling them with their respective feature recognition types; and analyzing the feature combination data belonging to the same feature recognition type using corresponding convolutional neural networks according to each feature recognition type to obtain the identified building.

[0007] For traditional Chinese architecture, not only the external outline of the building is collected, but also other parts of the building are collected as feature data. The collected feature data are preprocessed and combined according to category to form a single feature dataset. This synthesized data is then analyzed using a convolutional neural network to identify the building. In this way, by combining various features for identification, the building's specifications and age can be determined.

[0008] Optionally, the image type includes overall outline, support structure, and painted texture.

[0009] The main distinguishing features between traditional Chinese buildings and modern buildings are not only their external outlines, but also their supporting structures and painted textures. Furthermore, there are differences in the supporting structures and painted textures among different traditional Chinese buildings. Extracting these key distinguishing features facilitates feature recognition by subsequent neural networks.

[0010] Optionally, the feature image is processed using an appropriate method according to the image type, including: using a differential hashing algorithm to obtain a preprocessed image of the overall contour for the feature image of the overall contour; or using a differential hashing algorithm based on Radon transform to obtain a preprocessed image of the support structure for the feature image of the support structure; or using LBP (Local Binary Pattern) to obtain a preprocessed image of the painted texture for the feature image of the painted texture.

[0011] The overall outline of the building can be extracted using a differential hashing algorithm. Because the internal support structure of the building is photographed from different angles, the captured images will be rotated compared to the standard image. Applying a Radon transform-based differential hashing algorithm to the support structure images can resist image rotation, and the obtained features are unique. For the exterior paintwork, due to its age and corrosion from the external environment, LBP is used to extract only the texture features of the paintwork and perform feature recognition on the paintwork texture.

[0012] Optionally, the LBP is a rotationally invariant uniform mode LBP.

[0013] Because the angle at which the exterior paintings of buildings are photographed varies, the resulting images will differ from standard images. Using the rotation-invariant uniformity (LBP) mode can resist rotation, ensuring the uniqueness of the obtained features.

[0014] Optionally, the preprocessed images are combined into feature combination data, including: combining the preprocessed images in the order of overall contour and supporting structure into a first feature recognition type; and combining the preprocessed images in the order of overall contour, supporting structure, and painted texture into a second feature recognition type.

[0015] The preprocessed images are combined according to their image type, with different combinations exhibiting different features. Classifying these features facilitates subsequent convolutional neural network analysis.

[0016] Optionally, based on each feature recognition type, a corresponding convolutional neural network is used to analyze feature combination data belonging to the same feature recognition type, including: using a basic feature convolutional neural network to analyze feature combination data belonging to the first feature recognition type; and using a full feature convolutional neural network to analyze feature combination data belonging to the second feature recognition type.

[0017] The preprocessed images are combined according to their image type, and different combinations exhibit different features. Appropriate convolutional neural networks are then used to analyze these different features, which helps improve the accuracy of feature recognition.

[0018] Optionally, the basic feature convolutional neural network is a multi-channel convolutional neural network; or, the full feature convolutional neural network is a multi-channel convolutional neural network.

[0019] Since the preprocessed image is a combination of building outlines and features of multiple building parts, a multi-channel convolutional neural network is used to identify multiple features.

[0020] In some embodiments, the apparatus includes a processor, a communication interface, and a memory storing program instructions, the processor being configured to execute, when running the program instructions, the method for visual recognition of traditional Chinese architecture as described above.

[0021] In some embodiments, the server includes the device for visual recognition of traditional Chinese architecture as described above.

[0022] In some embodiments, the storage medium includes: storing program instructions, characterized in that, when the program instructions are executed, they perform the method for visual recognition of traditional Chinese architecture as described above.

[0023] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0024] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrations and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are shown as similar elements. The drawings are not to be scaled. And wherein:

[0025] Figure 1 It is a visual identity architecture diagram of the building;

[0026] Figure 2 This is an architecture diagram for visual recognition of traditional Chinese architecture provided in an embodiment of this disclosure;

[0027] Figure 3 This is a flowchart illustrating a method for visual recognition of traditional Chinese architecture provided in an embodiment of this disclosure.

[0028] Figure 4 This is a flowchart of a method for visual recognition of traditional Chinese architecture provided in an embodiment of this disclosure;

[0029] Figure 5This is a schematic diagram illustrating an example of visual recognition of traditional Chinese architecture provided in this embodiment of the disclosure;

[0030] Figure 6 This is a schematic diagram of another device for visual recognition of traditional Chinese architecture provided in this embodiment. Detailed Implementation

[0031] To provide a more detailed understanding of the features and technical content of the embodiments of this disclosure, the implementation of the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this disclosure. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0032] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this disclosure described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0033] Unless otherwise stated, the term "multiple" means two or more.

[0034] The term "correspondence" can refer to an association or binding relationship. The correspondence between A and B means that there is an association or binding relationship between A and B.

[0035] Combination Figure 1 This paper describes a building visual recognition architecture. Existing methods for building recognition using convolutional neural networks (CNNs) involve: first, acquiring an image of the building's external contour, using this contour as the primary feature data; then, using a differential hashing algorithm to extract features from the external contour data, extracting the overall contour image, removing color values ​​and retaining only grayscale values, and compressing the extracted contour into a fixed-pixel image as a preprocessed overall contour image; finally, inputting this preprocessed overall contour image into a convolutional neural network (CNN) to recognize the building.

[0036] Combination Figure 2 The present disclosure provides an architecture diagram for visual recognition of traditional Chinese architecture, including:

[0037] First, feature images were collected for specific parts of the traditional Chinese building. These images included the overall outline of the building, images of the internal supporting structure, and images of the painted textures on the exterior. Then, the image types were labeled: the overall outline image was labeled as "overall outline," the internal supporting structure image as "supporting structure," and the external painted texture image as "painted texture."

[0038] For feature images labeled as overall contours, a differential hashing algorithm is used for feature extraction and image compression to obtain a preprocessed image of the overall architecture. For feature images labeled supporting structures, a Radon transform-based differential hashing algorithm is used for rotation-resistant feature extraction and image compression to obtain a preprocessed image of the supporting structures. If the feature images contain painted images, the LBP algorithm is used for feature extraction and image compression of the painted images to obtain a preprocessed image of the painted texture.

[0039] To recognize multiple feature images using a multi-channel convolutional neural network, the features need to be combined and classified. The first feature recognition type includes preprocessed images of the overall contour and preprocessed images of supporting structures, combined to form the first feature combination data. The second feature recognition type includes preprocessed images of the overall architecture, preprocessed images of supporting structures, and preprocessed images of painted textures, combined to form the second feature combination data.

[0040] If only the first feature combination data is available, a first convolutional neural network is used to analyze the first feature combination data to obtain the identified building. If the second feature combination data is available, a second convolutional neural network is used to analyze the second feature combination data to obtain the identified building.

[0041] Combination Figure 3 As shown in the embodiments of this disclosure, a processing flow for a method of visual recognition of traditional Chinese architecture includes:

[0042] S01: Collect feature images of the building as a whole and multiple parts of the building by taking pictures, and label the image type of the feature images.

[0043] S02, the Chinese ancient architecture recognition software uses the corresponding method to preprocess the feature image according to the image type to obtain the preprocessed image.

[0044] S03, the Chinese ancient architecture recognition software combines each preprocessed image into feature combination data and labels them with their respective feature recognition types.

[0045] S04, the Chinese ancient building recognition software uses a corresponding convolutional neural network to analyze feature combination data belonging to the same feature recognition type according to each feature recognition type, and obtains the recognized building.

[0046] This disclosure provides a method for visual recognition of Chinese ancient buildings. It can combine various unique features of Chinese ancient buildings based on their external outline, internal support structure, and external paintings, and use convolutional neural networks to accurately distinguish the age and specifications of Chinese ancient buildings.

[0047] Optionally, the image type in step S01 includes overall outline, support structure, and painted texture.

[0048] In this way, because the overall outline, supporting structure, and painted textures of traditional Chinese architecture are significantly different from those of modern buildings, these features can be used to distinguish between ancient and modern buildings. Furthermore, the supporting structure and painted textures can also be used to differentiate the age and specifications of ancient buildings, facilitating accurate feature recognition by subsequent convolutional neural networks.

[0049] Optionally, step S02 includes processing the feature image using an appropriate method according to the image type, including: using a differential hashing algorithm to obtain a preprocessed image of the overall contour for the feature image of the overall contour; or using a differential hashing algorithm based on Radon transform to obtain a preprocessed image of the support structure for the feature image of the support structure; or using LBP to obtain a preprocessed image of the painted texture for the feature image of the painted texture.

[0050] Thus, the overall outline of the building can be extracted using a differential hashing algorithm. Because the building's supporting structure was photographed from different angles, the captured images will be rotated compared to the standard image. Processing the supporting structure image with a differential hashing algorithm based on Radon transform can resist image rotation, ensuring the uniqueness of the obtained features. For the building's exterior paintings, due to their age and environmental corrosion, LBP is used to extract only the texture features of the paintings, and feature recognition is performed on the painted texture.

[0051] Optionally, the LBP described above is a rotationally invariant uniform mode LBP.

[0052] Thus, because the angle at which the exterior paint on a building is photographed varies, the resulting image will differ from the standard image due to rotation. Using the rotation-invariant uniform mode (LBP) can resist rotation, ensuring the uniqueness of the obtained features.

[0053] Optionally, the step S03 of combining the preprocessed images into feature combination data includes: combining the preprocessed images of the overall contour and supporting structure in the order of the preprocessed images into a first feature recognition type; and combining the preprocessed images of the overall contour, supporting structure, and painted texture in the order of the preprocessed images into a second feature recognition type.

[0054] In this way, the preprocessed images are combined according to their image type, and different combinations exhibit different features. Classifying different features facilitates subsequent convolutional neural network analysis.

[0055] Optionally, step S04, which involves using a corresponding convolutional neural network to analyze feature combination data belonging to the same feature recognition type based on each feature recognition type, includes: using a basic feature convolutional neural network to analyze feature combination data belonging to the first feature recognition type; and using a full feature convolutional neural network to analyze feature combination data belonging to the second feature recognition type.

[0056] In this way, the preprocessed images are combined according to their image type, and different combinations exhibit different features. Appropriate convolutional neural networks are then used to analyze these different features, which helps improve the accuracy of feature recognition.

[0057] Optionally, the above-mentioned basic feature convolutional neural network is a multi-channel convolutional neural network; or, the full feature convolutional neural network is a multi-channel convolutional neural network.

[0058] Since the preprocessed image is a combination of building outlines and features of multiple building parts, using a multi-channel convolutional neural network to identify multiple features can improve the feature recognition rate.

[0059] Combination Figure 4 As shown in the figure, the flow of a method for visual recognition of traditional Chinese architecture provided in this disclosure includes:

[0060] S100: Collect the overall outline of the building's exterior.

[0061] S101, obtain a preprocessed image of the overall outline of the building through a differential hashing algorithm.

[0062] S102, Collect the internal support structure diagram of the building.

[0063] S103, a preprocessed image of the building support structure is obtained by using a differential hashing algorithm based on Radon transform.

[0064] S104. Does the exterior of the building have painted textures? If it does, proceed to S105; otherwise, proceed to S109.

[0065] S105, captures images of the painted textures on the exterior of buildings.

[0066] S106, obtain the preprocessed image of the painted texture through LBP.

[0067] S107, the preprocessed images of the overall outline, supporting structure, and painted texture are combined into the second feature combination data.

[0068] S108: The second feature combination data is analyzed using a full-feature convolutional neural network, and the process ends after identifying the building.

[0069] S109, the preprocessed images of the overall outline and supporting structure are combined into the first feature combination data.

[0070] S110: The first feature combination data is analyzed using a basic feature convolutional neural network, and the process ends after identifying the building.

[0071] Combination Figure 5 As shown, this disclosure provides an example of visual recognition for traditional Chinese architecture.

[0072] First, images of the overall external outline of the traditional Chinese building, images of the internal supporting structure, and images of the painted textures on the exterior were collected. Then, the image types were labeled: the overall outline image was labeled as "overall outline," the internal supporting structure image was labeled as "supporting structure," and the painted texture image was labeled as "painted texture."

[0073] For feature images labeled as overall contours, a differential hashing algorithm is used for feature extraction and image compression to obtain a preprocessed image of the overall architecture. For feature images labeled as supporting structures, a Radon transform-based differential hashing algorithm is used for rotation-resistant feature extraction and image compression to obtain a preprocessed image of the supporting structures. For feature images labeled as painted textures, the LBP algorithm is used for feature extraction and image compression to obtain a preprocessed image of the painted textures.

[0074] The preprocessed images of the overall architecture, the preprocessed images of the supporting structures, and the preprocessed images of the painted textures are combined into a second feature combination data.

[0075] The second feature combination data is input into a full-feature convolutional neural network. The full-feature convolutional neural network analyzes the second feature combination data to obtain the identified buildings.

[0076] Combination Figure 6 As shown, this disclosure provides an apparatus for visual recognition of traditional Chinese architecture, including a processor 100 and a memory 101. Optionally, the apparatus may further include a communication interface 102 and a bus 103. The processor 100, communication interface 102, and memory 101 can communicate with each other via the bus 103. The communication interface 102 can be used for information transmission. The processor 100 can call logical instructions in the memory 101 to execute the method for visual recognition of traditional Chinese architecture described in the above embodiment.

[0077] Furthermore, the logic instructions in the aforementioned memory 101 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0078] The memory 101, as a storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor 100 executes functional applications and data processing by running the program instructions / modules stored in the memory 101, thereby implementing the method for visual recognition of Chinese ancient buildings in the above embodiments.

[0079] The memory 101 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 101 may include high-speed random access memory and may also include non-volatile memory.

[0080] This disclosure provides a server that includes the aforementioned device for visual recognition of traditional Chinese architecture.

[0081] This disclosure provides a storage medium storing computer-executable instructions configured to execute the above-described method for visual recognition of traditional Chinese architecture.

[0082] The aforementioned storage medium can be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0083] The technical solutions of this disclosure can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in this disclosure. The aforementioned storage medium can be a non-transitory storage medium, including: a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and other media capable of storing program code; it can also be a transient storage medium.

[0084] The foregoing description and accompanying drawings fully illustrate embodiments of this disclosure to enable those skilled in the art to practice them. Other embodiments may include structural, logical, electrical, procedural, and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the terminology used in this application is for describing embodiments only and is not intended to limit the claims. As used in the description of embodiments and claims, the singular forms “a,” “an,” and “the” are intended to equally include the plural forms unless the context clearly indicates otherwise. Similarly, the term “and / or” as used in this application means including one or more of the associated listed items and all possible combinations thereof. Additionally, when used in this application, the term "comprise" and its variations "comprises" and / or "comprising" refer to the presence of stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. Without further limitations, an element defined by the phrase "comprises a..." does not exclude the presence of other identical elements in the process, method, or apparatus that includes said element. In this document, each embodiment may focus on the differences from other embodiments, and similar or identical parts between embodiments can be referred to mutually. For methods, products, etc., disclosed in the embodiments, if they correspond to the method section disclosed in the embodiments, the relevant parts can be referred to the description of the method section.

[0085] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this disclosure. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0086] The methods and products (including but not limited to apparatuses and devices) disclosed in the embodiments herein can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative.

[0087] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

Claims

1. A method for visual recognition of traditional Chinese architecture, characterized in that, include: Collect feature images of the building as a whole and multiple parts of the building, and label the image types of the feature images; The feature image is preprocessed according to the image type using the appropriate method to obtain the preprocessed image; The preprocessed images are combined into feature combination data, and their respective feature recognition types are labeled. Based on each feature recognition type, the corresponding convolutional neural network is used to analyze the feature combination data belonging to the same feature recognition type to obtain the specifications and age of the identified building; The image types include overall outline, supporting structure, and painted texture; According to the image type, the feature images are preprocessed using appropriate methods, including: for feature images of the overall contour, a differential hashing algorithm is used to obtain a preprocessed image of the overall contour; for feature images of the supporting structure, a differential hashing algorithm based on Radon transform is used to obtain a preprocessed image of the supporting structure; for feature images of painted textures, a Local Binary Pattern (LBP) is used to obtain a preprocessed image of the painted texture. The preprocessed images are combined into feature combination data, including: if there is no painted texture image, the preprocessed images are combined in the order of overall outline and supporting structure to form the first feature recognition type; if there is a painted texture image, the preprocessed images are combined in the order of overall outline, supporting structure and painted texture to form the second feature recognition type. Based on each feature recognition type, the corresponding convolutional neural network is used to analyze feature combination data belonging to the same feature recognition type, including: using a basic feature convolutional neural network to analyze feature combination data belonging to the first feature recognition type; and using a full feature convolutional neural network to analyze feature combination data belonging to the second feature recognition type.

2. The method according to claim 1, characterized in that, The LBP is a rotationally invariant uniform mode LBP.

3. The method according to claim 2, characterized in that, The basic feature convolutional neural network is a multi-channel convolutional neural network; or, the full feature convolutional neural network is a multi-channel convolutional neural network.

4. A device for visual recognition of traditional Chinese architecture, comprising a processor, a communication interface, and a memory storing program instructions, characterized in that, The processor is configured to execute, when running the program instructions, the method for visual recognition of Chinese ancient architecture as described in any one of claims 1 to 3.

5. A server, characterized in that, Includes the device for visual recognition of traditional Chinese architecture as described in claim 4.

6. A storage medium storing program instructions, characterized in that, When the program instructions are executed, they perform the method for visual recognition of Chinese ancient buildings as described in any one of claims 1 to 3.

Citation Information

Patent Citations

  • Historical building chronicle image measurement technology

    CN107273385A

  • Method for classifying a building, machine learning system and method, computer program and computer device for training the machine learning system

    EP3671560A1