WO2025102445A1 - 一种进行花纹演变追踪的可视分析方法及系统 - Google Patents

一种进行花纹演变追踪的可视分析方法及系统 Download PDF

Info

Publication number
WO2025102445A1
WO2025102445A1 PCT/CN2023/136321 CN2023136321W WO2025102445A1 WO 2025102445 A1 WO2025102445 A1 WO 2025102445A1 CN 2023136321 W CN2023136321 W CN 2023136321W WO 2025102445 A1 WO2025102445 A1 WO 2025102445A1
Authority
WO
WIPO (PCT)
Prior art keywords
pattern
evolution
patterns
tracking
visual analysis
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/CN2023/136321
Other languages
English (en)
French (fr)
Inventor
袁晓如
李金城
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Peking University
Original Assignee
Peking University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Peking University filed Critical Peking University
Publication of WO2025102445A1 publication Critical patent/WO2025102445A1/zh
Anticipated expiration legal-status Critical
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/54—Browsing; Visualisation therefor
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces

Definitions

  • the invention belongs to the field of pattern evolution analysis, and in particular relates to a visual analysis method and system for pattern evolution tracking.
  • Patterns are usually painted as decorations on artifacts such as utensils, clothing, and buildings. Patterns of the same period generally follow common norms and styles. In-depth analysis of pattern evolution helps reveal the characteristics of the spread and development of culture in different eras and regions. Therefore, pattern evolution analysis is often the primary and key method for archaeologists to study ancient Chinese painted pottery, clothing, architecture and other cultural relics.
  • the purpose of the present invention is to provide a visual analysis method and system for tracking pattern evolution, to establish a special analysis system that can intelligently track pattern evolution, and to provide an interface for users to organize the process and results of evolution tracking, thereby improving the convenience and efficiency of pattern evolution analysis, providing reliable support for archaeologists in the entire process of pattern evolution analysis, and laying a foundation for the development of cultural relics archaeology.
  • a visual analysis method for pattern evolution tracking comprising the following steps:
  • step S1 includes taking the appearance information, time information and space information of the pattern to be analyzed as input, generating a pattern overview with adjustable similarity weight measurement indicators of the pattern to be analyzed, selecting one or more patterns for search and matching according to the pattern overview of the pattern to be analyzed, and selecting patterns with evolutionary relationships according to different similarities of the patterns.
  • the similarity weight measurement index of the patterns in step S1 includes appearance similarity, time similarity and space similarity between the patterns.
  • the pattern overview in step S1 includes a pattern image, a time distribution, and a space distribution of the pattern to be analyzed.
  • step S2 the evolution order of the patterns is determined according to the appearance similarity, time attribute and space attribute of the selected patterns.
  • step S2 when the pattern evolution includes multiple different stages, the pattern is divided into several subgroups, each subgroup corresponding to a stage.
  • step S1 includes checking similar pattern results matched according to different similarities according to user input.
  • step S2 includes organizing the established pattern evolution sequence according to user input.
  • the pattern layout is displayed as a sequence-based layout and a space-based layout.
  • a visual analysis system for tracking pattern evolution uses a visual analysis method for tracking pattern evolution as described in the first aspect of the present invention and any optional embodiment thereof to track and analyze the evolution process of a pattern to be analyzed.
  • the beneficial technical effects of the present invention are: using a visual analysis method and system for pattern evolution tracking disclosed in the present invention, by selecting one or more patterns as analysis entries according to the characteristics of the pattern to be analyzed, and then searching and matching similar patterns according to the analysis entry, a group of patterns with evolutionary relationships are obtained, the selected patterns are organized, the evolution sequence of the selected patterns is determined, and whether to divide the pattern into subgroups is determined according to the degree of change; and the findings and insights in the pattern evolution analysis process are recorded.
  • a special analysis system which can intelligently track the evolution of patterns, visualize the evolution tracking process, and provide an interface for users to organize the process and results of evolution tracking, which can provide full-process support for experts to conduct pattern evolution analysis and research, significantly improve analysis efficiency, and lay a foundation for the development of cultural relics archaeology.
  • FIG1 is a flow chart of a visual analysis method for pattern evolution tracking disclosed in Embodiment 1 of the present invention.
  • FIG2 is a schematic diagram of a working process interface of a visual analysis system for pattern evolution tracking disclosed in Embodiment 2 of the present invention.
  • FIG3 is a diagram showing an interface of a pattern selection module in a visual analysis system for pattern evolution tracking disclosed in a second embodiment of the present invention
  • FIG4 is a diagram of an interface of a pattern search interaction process of a pattern selection module in a visual analysis system for pattern evolution tracking disclosed in a second embodiment of the present invention
  • FIG5 is a diagram showing an interface of a pattern organization module in a visual analysis system for pattern evolution tracking disclosed in Embodiment 2 of the present invention
  • FIG. 6 is an interface diagram of an evolution process recording module in a visual analysis system for pattern evolution tracking disclosed in Embodiment 2 of the present invention.
  • an embodiment of the present invention provides a visual analysis method for pattern evolution tracking, the method comprising the following steps:
  • Patterns with evolutionary relationships are usually similar in appearance, close in space, or adjacent in time.
  • the appearance information, time information, and space information of the pattern to be analyzed are used as inputs to generate a pattern overview with adjustable similarity weight measurement indicators of the pattern to be analyzed, and one or more patterns are selected for search and matching based on the pattern overview of the pattern to be analyzed, and patterns with evolutionary relationships are selected from them based on the appearance similarity, time relationship, and space relationship between the patterns.
  • the similarity weight measurement indicators of patterns include appearance similarity, time similarity and spatial similarity between patterns.
  • the specific values of the similarity weight measurement indicators are determined based on user input.
  • the pattern overview includes images of all patterns to be analyzed and their temporal and spatial distribution to help the user select one or more similar patterns.
  • Step S1 includes checking similar pattern results matched according to different similarities according to user input.
  • An evolution sequence is established based on the relationship between the selected patterns, and the patterns can be divided into different subgroups according to the degree of change.
  • Step S2 includes determining the evolution order of the patterns according to the appearance similarity and the time and space attributes of the selected patterns.
  • the pattern evolution includes multiple different stages
  • the pattern is divided into several subgroups, each subgroup corresponds to a stage. After the division is completed, the user can also add a description of each subgroup.
  • Step S2 includes organizing the established pattern evolution sequence according to user input.
  • the system Receive the pattern evolution process summary input by the user, record and store the user's observations and insights.
  • the system displays the pattern layout in two forms: sequence-based layout and space-based layout.
  • the system supports multiple annotation forms including text description, outlines for highlighting local patterns, and labels.
  • An embodiment of the present invention provides a visual analysis system for tracking pattern evolution.
  • the system uses a visual analysis method for tracking pattern evolution as described in Embodiment 1 of the present invention and any optional implementation manner thereof to track and analyze the evolution process of a pattern to be analyzed.
  • a visual analysis system for pattern evolution tracking disclosed in an embodiment of the present invention is used to track and analyze the evolution process of the pattern to be analyzed, with pattern information, time information and space information as system input.
  • the system includes a pattern selection module a, a pattern organization module b and an evolution process recording module c, wherein the pattern selection module a is used to select a group of patterns with evolution relationships, the pattern organization module b is used to determine the evolution sequence of the selected pattern and group the patterns, and the evolution process recording module c is used to record the findings and insights in the evolution process.
  • the pattern selection module includes the following five units.
  • Timeline view Used to show the time distribution of patterns. Solid rectangles represent types, and dashed rectangles represent cultures. Similar to the spatial tree view, each rectangle has a shaded ring with a number inside, indicating the proportion and number of patterns related to the corresponding culture or type.
  • Triprojection View It is used to integrate the appearance similarity of patterns with temporal and spatial proximity using a weighting mechanism.
  • the weights of the three parts are controlled by a triangular control to control. Users can modify the weights by moving the anchor points inside the control, and the weight corresponding to each vertex will increase as the anchor points get closer.
  • the appearance similarity weight is 1
  • the view shows the similarity of the pattern images. The similarity is obtained using the perception-based metric LPIPS, and then the pattern layout is obtained using t-SNE.
  • the spatial weight is 1, the view is displayed as an abstract map with the pattern image superimposed on its corresponding city.
  • the time weight is set to 1, the pattern image is aligned along the x-axis with the type center in the timeline view, and its y-axis position is determined by the appearance similarity weight and the spatial weight in the previous step.
  • Pattern Group View It is used to display the created pattern groups with evolution relationship. Each row represents a pattern group, and the columns indicate the pattern group name, subgroups and number of patterns. Users can delete, organize and enter the record interface through the three icons from left to right.
  • Diffusion lens Since the projected view may have visual occlusion of the pattern, this problem is alleviated by introducing a diffusion lens. After activation, the user can click on a specific area in the view that they want to take a closer look. After that, the pattern around the click position will automatically spread outward, and the diffusion direction is determined by the click position and the center of the image.
  • the interactive process of pattern search in the pattern selection module is detailed in the pie menu.
  • the pie menu is triggered by clicking on the pattern image. It contains two parts: filtering and creating, which are used to perform pattern-based search and pattern group operations respectively.
  • the filtering part is divided into three operations: similarity search Spatial Search and time search
  • the outermost layer of the filter is a number of concentric rings that act as parameter controllers. For similarity, they control the number of patterns to be displayed based on how similar the pattern is to the target pattern. For space and time, they determine the number of cities and types to be displayed that are close to the target pattern. All interactions in the pie menu are triggered by hovering the mouse. This interactive experience is smoother than clicking. Hovering over the concentric rings will also display text prompts next to the rings, explaining the specific action to be performed.
  • Similarity Search Considering that global projection distortion may affect the user's perception of the similarity between patterns, the system uses ego-centric projection to provide a more reliable similarity relationship between the target pattern and the patterns that meet the search criteria.
  • the target pattern is located in the center of the view and is distinguished by a solid brown outline.
  • the radial distance between other patterns and the target pattern represents the exact similarity between them.
  • the angular difference between other patterns represents the similarity between them.
  • the spatial search Compared to the global layout adjustment controlled by the triangle control, the spatial search only displays patterns in adjacent cities, solving the problem of overlapping results by saving screen space. Specifically, the system handles overlaps within cities and between cities. The system determines the arrangement of patterns within cities through circular packing technology. When dealing with overlaps between cities, the patterns in the affected cities are repositioned together according to the relative directions of the cities. Bubble Sets are used to provide clear boundaries for each city.
  • time Search Similar to spatial search, the optimization of temporal search results compared to directly setting the time weight to 1 through the triangle control is also mainly reflected in solving overlaps.
  • the initial layout of the results encodes the type of pattern through the x-direction position and the y-direction position encodes the similarity between patterns. In a type, if two patterns overlap, the pattern below will move downward. If the bottom pattern is blocked, the system will ensure the visibility of all patterns by adjusting their overall position, even if this will cause slight overlap.
  • the system uses a strategy similar to spatial search to handle them. The difference is that patterns are only allowed to move horizontally.
  • the interface of the pattern organization module b includes the following two parts:
  • the second icon in the dialog box is used to determine the evolution sequence of the pattern.
  • the triple projection only displays the previously selected pattern. Their positions along the x-axis indicate their type, and their positions along the y-axis indicate their similarity in appearance. Due to the introduction of overlapping strategies and intervals, the positions of the patterns in both directions can only roughly reflect their temporal and similarity relationships. Each pattern displays its temporal and spatial properties directly below its image. By clicking on two patterns with a direct evolutionary relationship, the user can build a line that represents the order of evolution.
  • the evolution of patterns sometimes manifests itself in different stages.
  • the system allows users to divide patterns into subgroups, each of which corresponds to a specific stage. Users can use the group operation list In this mode, the pattern layout remains the same as in the sequence building mode, and the subgroup operation panel appears.
  • the Add icon on the far right is used to create a new subgroup. Each new subgroup is assigned a unique color.
  • the system supports three operations: modify subgroup members, delete subgroup, and add description (three icons from left to right). When modifying members, users can add patterns to the subgroup by clicking the corresponding image.
  • the interface of the evolution process recording module c includes the following parts:
  • the spatial tree view and the timeline view also exist in the pattern selection module a interface. They are simplified to only display the spatial and temporal distribution of the selected pattern and display the image thumbnails of the pattern.
  • the Subgroups panel located at the top of the interface, is used to add descriptions for each subgroup.
  • a circular icon used to display the image, temporal and spatial attributes of each pattern, and user-provided information.
  • the color of the icon indicates the subgroup to which the pattern belongs, and the number inside indicates the order of the pattern in the evolution sequence. Arrows connect adjacent patterns in the evolution sequence.
  • the view provides sequence-based and space-based There are two layout modes, and users can switch between them through the radio button in the navigation bar at the top of the interface.
  • the pattern arrangement in the view is not like , panels showing spatiotemporal properties are represented by icons and are collapsed by default.
  • Sequence-based layout view The pattern follows a zigzag arrangement in this layout.
  • Space-based layout view The patterns are superimposed on a city-centered map. Each city is represented by its center and label.
  • the layout of the patterns in each city is determined by a force-directed strategy.
  • the strategy involves three forces: city center force, boundary force, and collision force. Together, they ensure that: (1) patterns within a city are close to each other; (2) patterns are mainly located within the city boundaries; (3) patterns do not overlap each other or with the city center and label.
  • the system sets an angle for the movement of each pattern based on its order. This linear mapping ensures that adjacent patterns in the sequence move in similar directions.
  • Annotation Panel Three types of annotations (text description, outlines of highlighted patterns, and labels) are provided to help users record their findings and insights. Labels include overall labels and local labels, which are used to describe the overall evolution trend and pairwise changes between patterns, respectively.
  • the system adopts a semi-automatic strategy to bind and add annotations. Specifically, it can automatically identify the visual elements bound to the user's annotations. Descriptions and labels are bound to the lines connecting adjacent patterns. They are then automatically added to the starting position of the corresponding lines in the other layout. Users can drag and adjust as needed. The outline is bound to the closest pattern. Given that the size of patterns in different layouts may vary, the system calculates the position and size of the newly added outline in the other layout based on its relative position and size to the pattern in the original layout.
  • the visual analysis method and system for pattern evolution tracking disclosed in the present invention are applied to the archaeology of cultural relics such as utensils, clothing and buildings.
  • One or more patterns are selected as analysis entries according to the characteristics of the pattern to be analyzed, and then similar patterns are searched and matched according to the analysis entry to obtain a group of patterns with evolutionary relationships.
  • the selected patterns are organized to determine the evolution sequence of the selected patterns, and whether to divide the pattern into subgroups according to the degree of change; the findings and insights in the pattern evolution analysis process are recorded.
  • the method disclosed in the present invention is used to establish a special analysis system that can intelligently track the evolution of patterns, and provide an interface for users to organize the process and results of evolution tracking, which can provide full-process support for experts to conduct pattern evolution analysis and research, significantly improve analysis efficiency, and lay a foundation for the development of cultural relics archaeology.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Library & Information Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Multimedia (AREA)
  • Processing Or Creating Images (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本发明涉及一种进行花纹演变追踪的可视分析方法及系统，属于花纹演变分析领域，通过根据待分析花纹特点选择一个或多个花纹作为分析入口，然后根据分析入口进行相似花纹搜索匹配，得到一组具有演变关系的花纹，对选中的花纹进行组织，确定选中花纹的演变序列，并根据变化程度确定是否划分花纹子组；记录花纹演变分析过程中的发现和见解。采用本发明中公开的方法，建立专门的分析系统，能够智能化地进行花纹演变追踪，并提供界面供用户对演变追踪的过程和结果进行组织，能够为专家进行花纹演变分析研究提供全流程支持，显著提升分析效率，为文物考古的发展奠定基础。

Description

一种进行花纹演变追踪的可视分析方法及系统 技术领域
本发明属于花纹演变分析领域，具体涉及一种进行花纹演变追踪的可视分析方法及系统。
背景技术
花纹通常作为装饰绘制在器具、服饰以及建筑物等文物上，同一时期的花纹一般遵循共同的规范和风格，对花纹演变的深入分析有助于揭示文化在不同时代和地区的传播和发展特征。因此，花纹演变分析往往是考古学家研究中国古代彩陶、服饰以及建筑等文物的首要和关键手段。
然而现有技术中没有专门针对花纹演变进行分析的工具，考古学家不得不手动从数以百计的花纹中筛选出具有演变关系的相同主题花纹，完成筛选后，他们只能通过文字来记录整体演变趋势和成对变化，整个分析过程完全依靠人工进行，非常繁琐费时。
发明内容
针对现有技术中存在的缺陷，本发明的目的在于提供一种进行花纹演变追踪的可视分析方法及系统，建立专门的分析系统，能够智能化地进行花纹演变追踪，并提供界面供用户对演变追踪的过程和结果进行组织，提高花纹演变分析的便捷性和效率，为考古学家进行花纹演变分析全过程提供可靠支持，为文物考古的发展奠定基础。
为达到以上目的，本发明采用的技术方案是：
第一方面，一种进行花纹演变追踪的可视分析方法，所述方法包括以下步骤：
S1、根据待分析花纹特点选择一个或多个花纹作为分析入口，然后根据分析入口进行相似花纹匹配，选中具有演变关系的完整花纹集合；
S2、对选中的花纹进行组织，以建立花纹演变序列和演变阶段；
S3、接收并保存用户通过多种形式输入的注释记录。
进一步，步骤S1中包括以待分析花纹的外观信息、时间信息和空间信息作为输入，生成待分析花纹相似性权重衡量指标可调节的花纹概览，根据待分析花纹的花纹概览选择一个或多个花纹进行搜索匹配，根据花纹的不同相似性选中具有演变关系的花纹。
进一步，步骤S1中花纹的相似性权重衡量指标包括花纹间的外观相似性和时间相似性以及空间相似性。
进一步，步骤S1中花纹概览包括待分析花纹的花纹图像、时间分布以及空间分布。
进一步，步骤S2中根据选中花纹的外观相似度、时间属性和空间属性，确定花纹的演变顺序。
进一步，步骤S2中当花纹演变包含多个不同的阶段时，将花纹划分为若干子组，每个子组与一个阶段对应。
进一步，步骤S1包括根据用户输入对按照不同相似性匹配的相似花纹结果进行查看。
进一步，步骤S2包括根据用户输入对建立的花纹演变序列进行组织。
进一步，步骤S3中用户输入注释记录时，花纹布局展示为基于序列的布局和基于空间的布局。
第二方面，一种进行花纹演变追踪的可视分析系统，采用如本发明第一方面及其任一可选实施方式所述的一种进行花纹演变追踪的可视分析方法对待分析的花纹进行演变过程追踪分析。
本发明的有益技术效果在于：采用本发明所公开的一种进行花纹演变追踪的可视分析方法及系统，通过根据待分析花纹特点选择一个或多个花纹作为分析入口，然后根据分析入口进行相似花纹搜索匹配，得到一组具有演变关系的花纹，对选中的花纹进行组织，确定选中花纹的演变序列，并根据变化程度确定是否划分花纹子组；记录花纹演变分析过程中的发现和见解。采用本发明中公开的方法，建立专门的分析系统，能够智能化地进行花纹演变追踪，并且将演变追踪过程进行可视化显示，并提供界面供用户对演变追踪的过程和结果进行组织，能够为专家进行花纹演变分析研究提供全流程支持，显著提升分析效率，为文物考古的发展奠定基础。
附图说明
图1为本发明实施例一中公开的一种进行花纹演变追踪的可视分析方法流程图；
图2为本发明实施例二中公开的一种进行花纹演变追踪的可视分析系统的工作过程界面示意图；
图3为本发明实施例二中公开的一种进行花纹演变追踪的可视分析系统中的花纹选择模块界面图；
图4为本发明实施例二中公开的一种进行花纹演变追踪的可视分析系统中的花纹选择模块的花纹搜索交互过程界面图；
图5为本发明实施例二中公开的一种进行花纹演变追踪的可视分析系统中的花纹组织模块界面图；
图6为本发明实施例二中公开的一种进行花纹演变追踪的可视分析系统中的演变过程记录模块界面图。
具体实施方式
下面结合附图和具体实施方式对本发明作进一步描述。
实施例一
如图1所示，本发明实施例提供一种进行花纹演变追踪的可视分析方法，所述方法包括以下步骤：
S1、根据待分析花纹特点选择一个或多个花纹作为分析入口，然后根据分析入口进行相似花纹匹配，选中具有演变关系的完整花纹集合。
存在演变关系的花纹通常在外观上相似、在空间上接近或者在时间上相邻。在本实施例中，以待分析花纹的外观信息、时间信息和空间信息作为输入，生成待分析花纹相似性权重衡量指标可调节的花纹概览，根据待分析花纹的花纹概览选择一个或多个花纹进行搜索匹配，根据花纹间的外观相似性和时间关系以及空间关系从中选中具有演变关系的花纹。
花纹的相似性权重衡量指标包括花纹间的外观相似性和时间相似性以及空间相似性,相似性权重衡量指标的具体数值根据用户输入进行确定。
花纹概览包括待分析花纹的所有花纹的图像及其时间和空间分布，以帮助用户选择一个或多个相似的花纹。
步骤S1包括根据用户输入对按照不同相似性匹配的相似花纹结果进行查看。
S2、对选中的花纹进行组织，以建立花纹演变序列和演变阶段。
根据选中花纹之间的关系建立演变序列，并可根据变化程度将花纹划分为不同的子组。
步骤S2包括根据选中花纹的外观相似度以及时间和空间属性，确定花纹的演变顺序。
当花纹演变包含多个不同的阶段时，将花纹划分为若干子组，每个子组与一个阶段对应。划分完成后，用户还可以添加对每个子组的描述。
步骤S2包括根据用户输入对建立的花纹演变序列进行组织。
S3、完成花纹选择和组织后，接收并保存用户通过多种形式的注释记录自己的观察和见解。
接收用户输入的花纹演变过程总结，记录并存储用户的观察与见解。为了支持用户灵活地进行记录，系统将花纹布局展示为基于序列的布局和基于空间的布局两种展示形式，系统支持包括文字描述、用于高亮局部花纹的轮廓以及标签等多种注释形式。
实施例二
本发明实施例提供一种进行花纹演变追踪的可视分析系统，所述系统采用如本发明实施例一及其任一可选实施方式所述的一种进行花纹演变追踪的可视分析方法对待分析的花纹进行演变过程追踪分析。
如图2所示，采用本发明实施例中公开的一种进行花纹演变追踪的可视分析系统对待分析的花纹进行演变过程追踪分析，以花纹信息、时间信息和空间信息作为系统输入。所述系统包括花纹选择模块a、花纹组织模块b以及演变过程记录模块c，花纹选择模块a用于选择一组具有演变关系的花纹,花纹组织模块b用于确定所选花纹的演变序列并对花纹分组，演变过程记录模块c用于记录演变过程中的发现和见解。
如图3所示，花纹选择模块包含以下五个单元。
其中空间树视图用于展示花纹的空间分布。其中省市通过矩形进行表示，矩形采用水平缩进的方式来区分省和市。每个矩形都包含一个名称标签，以及一个由阴影环和内部数字组成的图标。该图标表示与该省市相关的花纹比例和总数。为确保数据的可见性，每个城市内显示的比例是基于其所属省份的总数，而不是全部花纹。
时间线视图用于显示花纹的时间分布。实线矩形表示类型，虚线矩形表示文化。与空间树视图类似，每个矩形中都有一个阴影环，环内有一个数字，表示与相应文化或类型相关的花纹比例和数量。
三重投影视图用于采用权重机制集成花纹的外观相似性与时间和空间相近性。三个部分的权重通过三角控件进行控制。用户可以通过移动控件内部的锚点来修改权重，每个顶点对应的权重会随着锚点的接近而增加。当外观相似性权重为1时，视图会显示花纹图像的相似性。相似性通过基于感知的度量方法LPIPS获取，之后通过t-SNE获得花纹布局。当空间权重为1时，视图显示为一张抽象地图，花纹图像叠加在其对应的城市上。将时间权重设为1时，花纹图像会沿着x轴与时间轴视图中的类型中心对齐，其y轴位置由上一步中外观相似度权重和空间权重共同决定。
花纹组视图用于显示已创建的存在演变关系的花纹组。其中，每行表示一个花纹组，列指示花纹组名称、包含的子组和花纹数量。用户可以通过从左至右三个图标完成删除、组织以及进入记录界面的操作。
扩散透镜由于投影视图可能存在花纹视觉遮挡，通过引入扩散透镜来缓解这一问题。激活后，用户可以点击视图中想要仔细查看的特定区域。之后，点击位置周围的花纹会自动向外扩散，扩散方向由点击位置和图像中心决定。
如图4所示，花纹选择模块中的花纹搜索交互过程详见饼状菜单饼状菜单通过点击花纹图片触发，它包含筛选和创建两个部分，分别用于执行基于花纹的搜索和花纹组操作。筛选部分又分为三种操作：相似度搜索空间搜索和时间搜索筛选的最外层是一些作为参数控制器的同心圆环。在相似度方面，它们根据花纹与目标花纹的相似度来控制要显示的花纹数量。对于空间和时间，它们决定了与目标花纹接近的城市和类型的显示数量。饼状菜单中的所有交互都是由鼠标悬停触发。与点击相比，这种交互体验更加流畅。将鼠标悬停在同心圆环上还会在圆环旁边显示文字提示，说明要执行的具体操作。
相似度搜索考虑到全局投影畸变可能影响用户对花纹间相似度的感知，系统采用以自我为中心的投影(ego-centric projection)来提供目标花纹与符合搜索标准的花纹之间更可信的相似性关系。目标花纹位于视图的中心位置，以棕色实心轮廓加以区分。其他花纹与目标花纹的径向距离表示它们之间精确的相似度。此外，其他花纹之间的角度差异代表了它们之间的相似性。这些相似性通过全局优化方法模拟退火(simulated annealing)计算得出。
空间搜索相比于由三角控件控制的全局布局调整，空间搜索只展示处于相邻城市内的花纹，通过节省的屏幕空间解决结果中的重叠问题。具体来说，系统处理城市内部和城市之间的重叠。系统在在城市内部通过圆形填充(circular packing)技术来确定花纹的排布。在处理城市间重叠时，受影响城市中的花纹会根据城市的相对方向共同重新定位。Bubble Sets用于为每个城市提供清晰的边界。
时间搜索与空间搜索类似，时间搜索结果相比通过三角控件直接设置时间权重为1的优化也主要体现在解决重叠上。结果的初始布局通过x方向位置编码花纹所处类型，y方向位置编码花纹间相似度。在一个类型中，如果两花纹重叠，那么下方的花纹会向下移动。如果底部的花纹被遮挡，系统会通过调整它们的整体位置来确保所有花纹的可见性，即使这样会带来轻微的重叠。同时，对于类型间的重叠，系统采用类似于空间搜索的策略进行处理。区别在于花纹只允许水平移动。
如图5所示，花纹组织模块b的界面包括以下两个部分：
通过点击组操作列表中的第二个图标进入，用于确定花纹的演变序列(即演变顺序)。在该模式下，三重投影只显示之前选中的花纹它们沿x轴的位置表示所处的类型，沿y轴的位置表示外观上的相似性。由于引入了处理重叠策略和间隔，花纹两个方向上的位置只能大致反映它们的时间和相似度关系。每个花纹都在其图像下方直接展示时空属性。通过点击两个具有直接演变关系的花纹，用户可以构建一条表示演变顺序的连线。
花纹的演变有时会表现为不同的阶段。系统支持用户将花纹划分为子组，每个子组与一个特定阶段对应。用户可以通过组操作列表中的第三个图标访问该功能。在该模式下，花纹布局与序列构建模式保持一致，并出现子组操作面板最右侧的添加图标用于创建新子组。每个新子组都会被分配一个独特的颜色。对于每个子组，系统支持三种操作：修改子组成员、删除子组和添加描述(从左到右的三个图标)。修改成员时，用户可以通过点击相应的图像为子组添加花纹。
如图6所示，演变过程记录模块c界面包括以下部分：
空间树视图和时间线视图，这两个视图也存在于花纹选择模块a界面，被简化为只展示选中花纹的空间和时间分布，并展示花纹的图像缩略图。
位于界面顶部的子组面板，子组面板用于为每个子组添加的描述。
记录视图和用于展示每个花纹的图像、时间属性和空间属性以及展示用户提供信息的圆形图标。图标颜色表示花纹所属的子组，内部数字表示花纹在演变序列中的顺序。带箭头的连线将演变序列中相邻的花纹连接起来。视图提供基于序列和基于空间两种布局方式，用户可以通过界面顶部导航栏中的单选框进行切换。由于视图中的花纹排列不像视图中那样结构化，显示时空属性的面板用图标表示，并默认折叠。
其中基于序列的布局视图花纹在该布局下遵循之字形排列。基于空间的布局视图将花纹叠加在以城市为中心的地图上。每个城市由其中心和标签代表。每个城市的花纹通过力导向策略确定布局。该策略涉及三种力：城市中心点力、边界力和碰撞力。它们共同确保：(1)城市内的花纹相互靠近；(2)花纹主要位于城市边界内；(3)花纹既不会相互重叠，也不会与城市中心和标签重叠。此外，为了与演化序列保持空间上的一致性，系统根据每个花纹的顺序为其移动设定角度。这种线性映射确保了序列中相邻的花纹在移动方向上相似。
标注面板提供三种标注形式(文字描述、高亮花纹局部的轮廓、标签)来帮助用户记录他们的发现和见解。标签包含整体标签和局部标签两种，分别用于描述整体演变趋势以及花纹间的成对变化。为了减少标注两种布局所需的劳动，系统采用半自动策略来绑定和添加标注。具体来说，它能自动识别与用户标注绑定的视觉元素。描述和标签会绑定到连接相邻花纹的连线。然后它们会自动加入到另一个布局中相应连线的起始位置。用户可以根据需要拖动和调整。轮廓与最接近的花纹绑定。鉴于不同布局中的花纹尺寸可能会有所不同，系统会根据新添加的轮廓在原始布局中与花纹的相对位置和尺寸，计算其在另一布局中的位置和尺寸。
通过上述实施例可以看出，本发明公开的一种进行花纹演变追踪的可视分析方法及系统，应用于器具、服饰以及建筑物等文物考古中，通过根据待分析花纹特点选择一个或多个花纹作为分析入口，然后根据分析入口进行相似花纹搜索匹配，得到一组具有演变关系的花纹，对选中的花纹进行组织，确定选中花纹的演变序列，并根据变化程度确定是否划分花纹子组；记录花纹演变分析过程中的发现和见解。采用本发明中公开的方法，建立专门的分析系统，能够智能化地进行花纹演变追踪，并提供界面供用户对演变追踪的过程和结果进行组织，能够为专家进行花纹演变分析研究提供全流程支持，显著提升分析效率，为文物考古的发展奠定基础。
本发明所述的方法及系统并不限于具体实施方式中所述的实施例，本领域技术人员根据本发明的技术方案得出其他的实施方式，同样属于本发明的技术创新范围。

Claims (10)

  1. 一种进行花纹演变追踪的可视分析方法，所述方法包括以下步骤：
    S1、根据待分析花纹特点选择一个或多个花纹作为分析入口，然后根据分析入口进行相似花纹匹配，选中具有演变关系的完整花纹集合；
    S2、对选中的花纹进行组织，以建立花纹演变序列和演变阶段；
    S3、接收并保存用户通过多种形式输入的注释记录。
  2. 如权利要求1所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S1中包括以待分析花纹的外观信息、时间信息和空间信息作为输入，生成待分析花纹相似性权重衡量指标可调节的花纹概览，根据待分析花纹的花纹概览选择一个或多个花纹进行搜索匹配，根据花纹的不同相似性选中具有演变关系的花纹。
  3. 如权利要求2所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S1中花纹的相似性权重衡量指标包括花纹间的外观相似性和时间相似性以及空间相似性。
  4. 如权利要求2所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S1中花纹概览包括待分析花纹的花纹图像、时间分布以及空间分布。
  5. 如权利要求1所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S2中根据选中花纹的外观相似度、时间属性和空间属性，确定花纹的演变顺序。
  6. 如权利要求5所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S2中当花纹演变包含多个不同的阶段时，将花纹划分为若干子组，每个子组与一个阶段对应。
  7. 如权利要求1所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S1包括根据用户输入对按照不同相似性匹配的相似花纹结果进行查看。
  8. 如权利要求1所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S2包括根据用户输入对建立的花纹演变序列进行组织。
  9. 如权利要求1所述的一种进行花纹演变追踪的可视分析方法，其特征在于：步骤S3中用户输入注释记录时，花纹布局展示为基于序列的布局和基于空间的布局。
  10. 一种进行花纹演变追踪的可视分析系统，其特征在于：采用如权利要求1-9中的任一权利要求所述的一种进行花纹演变追踪的可视分析方法对待分析的花纹进行演变过程追踪分析。
PCT/CN2023/136321 2023-11-16 2023-12-05 一种进行花纹演变追踪的可视分析方法及系统 Pending WO2025102445A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202311528498.4 2023-11-16
CN202311528498.4A CN117786150A (zh) 2023-11-16 2023-11-16 一种进行花纹演变追踪的可视分析方法及系统

Publications (1)

Publication Number Publication Date
WO2025102445A1 true WO2025102445A1 (zh) 2025-05-22

Family

ID=90382222

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/136321 Pending WO2025102445A1 (zh) 2023-11-16 2023-12-05 一种进行花纹演变追踪的可视分析方法及系统

Country Status (2)

Country Link
CN (1) CN117786150A (zh)
WO (1) WO2025102445A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN121582906A (zh) * 2026-01-27 2026-02-27 安徽瑞轩供应链科技有限公司 基于人工智能的影像资料验证系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104484326A (zh) * 2014-09-30 2015-04-01 天津大学 一种基于可视分析的文物集成信息的交互探索方法
CN114187178A (zh) * 2021-12-13 2022-03-15 浙大城市学院 一种文物修复辅助用瓷器碎片分类、拼接系统及方法
CN116795997A (zh) * 2023-05-22 2023-09-22 北京邮电大学 一种面向文物领域的知识图谱构建方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104484326A (zh) * 2014-09-30 2015-04-01 天津大学 一种基于可视分析的文物集成信息的交互探索方法
CN114187178A (zh) * 2021-12-13 2022-03-15 浙大城市学院 一种文物修复辅助用瓷器碎片分类、拼接系统及方法
CN116795997A (zh) * 2023-05-22 2023-09-22 北京邮电大学 一种面向文物领域的知识图谱构建方法

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
AN NING: "Research on Evolutionary Relationship Discovery and Visualization of a Class of Traditional Decoration", CHINESE MASTER'S THESES FULL-TEXT DATABASE, 27 May 2021 (2021-05-27), pages 1 - 72, XP093316533, DOI: 10.26969/d.cnki.gbydu.2020.002213 *
LU MING , HOU XIAOGANG ,HAN XIAOTONG ,ZHAO HAIYING: "Discovery of Evolutionary Relationship of Ornaments Based on Knowledge Reasoning", COMPUTER ENGINEERING AND APPLICATIONS, vol. 58, no. 14, 15 July 2022 (2022-07-15), CN , pages 194 - 199, XP093316534, ISSN: 1002-8331, DOI: 10.3778/j.issn.1002-8331.2011-0118 *
ZHENG RUIZHENG: "Research on the Visualization of Plant Pattern Information in Yungang Grottoes", CHINESE MASTER'S THESES FULL-TEXT DATABASE, 1 June 2023 (2023-06-01), pages 1 - 44, XP093316535, DOI: 10.27284/d.cnki.gsxiu.2023.002280 *

Also Published As

Publication number Publication date
CN117786150A (zh) 2024-03-29

Similar Documents

Publication Publication Date Title
US6842176B2 (en) Computer-related method and system for controlling data visualization in external dimension(s)
US20180253204A1 (en) Configuring and displaying multidimensional data using two or more correlated interactive screen interfaces
US6628279B1 (en) System and method for three-dimensional modeling
US20100011309A1 (en) Data visualisation systems
US8321475B2 (en) System and method for contextual data modeling utilizing tags
US20090319948A1 (en) Automated editing of graphics charts
JPH08249352A (ja) 画像検索方法及び装置
Li et al. PM-Vis: A visual analytics system for tracing and analyzing the evolution of pottery motifs
Prakljačić et al. Tool for image annotation in context of modern object detection
CN117786150A (zh) 一种进行花纹演变追踪的可视分析方法及系统
JPH0830674A (ja) デザイン仕様決定支援システム
JPH076076A (ja) ハイパーメディアシステム
Baker et al. Interaction techniques for comparing video
US11263791B2 (en) System and method for generation of an interactive color workspace
JP3972692B2 (ja) 複数要素一括選択プログラムおよび複数要素一括選択方法
CN115509665A (zh) 一种录制窗口中控件的方法、装置、介质及设备
Agrawal et al. Hololabel: Augmented reality user-in-the-loop online annotation tool for as-is building information
JP4114407B2 (ja) 関係表示装置
JP2000123029A (ja) データ管理方法および装置、ならびにデータ管理方法をコンピュータで実行させるためのデータ管理プログラムを記録したコンピュータで読取り可能な記録媒体
Girgin et al. A streamlined GIS interface for Citizen Science activities: QGIS Light
JP2009026160A (ja) 階層データ表示プログラム，階層データ表示方法，階層データ表示装置
Bonerath et al. Algorithms for Consistent Dynamic Labeling of Maps With a Time-Slider Interface
JP2749041B2 (ja) データベースの編集方法
CN114141312A (zh) 一种基于vr的药效团虚拟筛选系统及方法
JP4212396B2 (ja) 図形要素選択プログラム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23958714

Country of ref document: EP

Kind code of ref document: A1