Intelligent tracking and checking method and system for museum collections and medium

Abstract

The invention discloses an intelligent museum collection tracking and checking method and system and a medium, and relates to the technical field of museum collection checking, and the method comprises the steps: collecting multi-dimensional data to generate a feature vector, and building a collection dynamic identity fingerprint; taking the collections, the exhibition stands and the fixing components in the exhibition space as nodes, and constructing an exhibition space constraint graph; microenvironment evolution data are collected; performing conjoint analysis on the dynamic identity fingerprint, the constraint graph and the microenvironment data, and calculating a residual error to obtain a consistency score; accordingly, triggering intelligent checking operation; and reporting an abnormal risk in combination with an operation result and the score. The technical problems that traditional museum collection checking depends on low manual inspection efficiency, environment data and collection states cannot be dynamically associated, and real-time risk early warning is difficult to realize are solved, and intelligent collection consistency evaluation is realized through multi-dimensional feature fusion, spatial constraint modeling and microenvironment evolution analysis, so that the collection checking efficiency is improved. The inventory efficiency and accuracy are improved, and the technical effect of whole-process intelligent management of collection protection is enhanced.

Description

Technical Field

[0001] This invention relates to the field of museum collection inventory technology, and in particular to intelligent tracking and inventory methods, systems and media for museum collections. Background Technology

[0002] In the daily operation and long-term preservation of museums, the inventory management of collections is a fundamental and crucial task. Traditional inventory methods mainly rely on regular manual inspections, counting, and recording, which has inherent drawbacks such as low efficiency, high labor costs, and susceptibility to human error. More importantly, faced with a large number and diverse range of precious collections, traditional methods cannot achieve real-time, continuous, and multi-dimensional monitoring of the physical condition, location security, and microenvironment of the collections. During exhibitions or storage, collections may experience subtle deformations, displacements, surface deterioration, or security risks that are difficult to detect with the naked eye due to environmental fluctuations, accidental touches, malicious acts, or their own aging. These potential anomalies are often only discovered during periodic manual inventory checks, missing the optimal intervention time and potentially leading to irreparable losses. Summary of the Invention

[0003] This invention provides a method, system, and medium for intelligent tracking and inventory of museum collections. It addresses the technical problems of traditional museum collection inventory relying on manual inspections, which are inefficient, unable to dynamically correlate environmental data with collection status, and difficult to achieve real-time risk warnings. The invention achieves the technical effect of intelligently assessing the consistency of collections through multi-dimensional feature fusion, spatial constraint modeling, and micro-environment evolution analysis, thereby improving inventory efficiency and accuracy and strengthening the intelligent management of the entire process of collection protection.

[0004] In a first aspect, the present invention provides a method for intelligent tracking and inventorying museum collections, wherein the method includes: The system collects 3D geometric contours, surface textures, and material response data of the collections to generate multi-dimensional identity feature vectors. These vectors are then used to establish dynamic identity fingerprints for the collections. Collections, display stands, and fixed components within the exhibition space are treated as nodes, and edges are established using relative position constraints, accessibility constraints, and synchronous change constraints between nodes to construct an exhibition space constraint graph. Microenvironmental data of the collections is collected to establish microenvironmental evolution data, including temperature, humidity, light, and vibration data. The dynamic identity fingerprints, exhibition space constraint graph, and microenvironmental evolution data are jointly analyzed to calculate residuals with historical data and associated collections, forming a collection consistency score. An inventory strategy is triggered based on the collection consistency score, and an intelligent inventory operation is performed. The results of the intelligent inventory operation and the collection consistency score are used to report any abnormal risks associated with the collections.

[0005] Secondly, the present invention also provides an intelligent tracking and inventory system for museum collections, wherein the intelligent tracking and inventory system for museum collections includes: The fingerprinting module collects the 3D geometric contours, surface textures, and material response data of the collections to generate multi-dimensional identity feature vectors. These vectors are then used to establish a dynamic identity fingerprint for each collection. The constraint construction module uses the collections, display stands, and fixed components in the exhibition space as nodes, and establishes edges based on relative position constraints, accessibility constraints, and synchronous change constraints between nodes, thus constructing an exhibition space constraint graph. The data acquisition module collects micro-environmental data from the collections, establishing micro-environmental evolution data including temperature, humidity, light, and vibration data. The joint analysis module performs joint analysis on the dynamic identity fingerprints, exhibition space constraint graph, and micro-environmental evolution data, calculating residuals with historical data and related collections to form a collection consistency score. The intelligent inventory module triggers an inventory strategy based on the collection consistency score and executes an intelligent inventory operation. The risk reporting module uses the results of the intelligent inventory operation and the collection consistency score to report abnormal risks associated with the collections.

[0006] Thirdly, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the intelligent tracking and inventory method for museum collections provided by the present invention.

[0007] This invention discloses a method, system, and medium for intelligent tracking and inventory of museum collections, including: collecting three-dimensional geometric contours, surface textures, and material response data of the collections to generate multi-dimensional identity feature vectors; establishing dynamic identity fingerprints of the collections using the multi-dimensional identity feature vectors; constructing an exhibition space constraint graph by using collections, display stands, and fixed components in the exhibition space as nodes and establishing edges based on relative position constraints, accessibility constraints, and synchronous change constraints between nodes; collecting micro-environmental data of the collections to establish micro-environmental evolution data including temperature, humidity, light, and vibration data; and jointly analyzing the dynamic identity fingerprints of the collections, the exhibition space constraint graph, and the micro-environmental evolution data to calculate and compare with historical data. The residuals of related collections are used to form a collection consistency score; the collection consistency score triggers an inventory strategy and executes intelligent inventory operations; the results of the intelligent inventory operations and the collection consistency score are used to report collection anomaly risks. The intelligent tracking and inventory method, system and medium for museum collections disclosed in this invention solves the technical problems of low efficiency, inability to dynamically associate environmental data and collection status, and difficulty in achieving real-time risk warning in traditional museum collection inventory. It achieves the technical effect of intelligent assessment of collection consistency through multi-dimensional feature fusion, spatial constraint modeling and micro-environment evolution analysis, improving inventory efficiency and accuracy, and strengthening the intelligent management of the entire process of collection protection. Attached Figure Description

[0008] Figure 1 This is a flowchart illustrating the intelligent tracking and inventory method for museum collections according to the present invention.

[0009] Figure 2 This is a schematic diagram of the intelligent tracking and inventory system for museum collections of the present invention.

[0010] Figure labeling: Fingerprint establishment module 11, constraint construction module 12, data acquisition module 13, joint analysis module 14, intelligent inventory module 15, risk reporting module 16. Detailed Implementation

[0011] The above technical solutions will now be described in detail with reference to the accompanying drawings and specific embodiments to provide a better understanding of them. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments used only to explain the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Furthermore, it should be noted that, for ease of description, only the parts related to the present invention are shown in the drawings, not all of them.

[0012] Example 1, as Figure 1 This is a flowchart illustrating the intelligent tracking and inventory method for museum collections according to the present invention, wherein the intelligent tracking and inventory method for museum collections includes: Collect the three-dimensional geometric contour, surface texture and material response data of the collection, generate a multi-dimensional identity feature vector, and use the multi-dimensional identity feature vector to establish a dynamic identity fingerprint of the collection.

[0013] Specifically, in the intelligent inventory process of museum collections, the external form of the collections is first scanned in three dimensions using 3D spatial sensing equipment, such as structured light 3D scanners and laser scanners, generating 3D geometric data that accurately reflects the volume, shape, size, and surface structure of the collections. Next, high-resolution imaging equipment, such as industrial-grade high-definition cameras and multispectral or hyperspectral imaging equipment, is used to collect information on the surface texture of the collections, obtaining information such as the material's gloss, texture details, and color variations. This data helps identify and distinguish different collections and provides more clues about the material and age of the collections. Simultaneously, sensors for temperature, humidity, light, and vibration, as well as non-contact detection devices for sensing the electrical, optical, or mechanical responses of materials, are used to collect material response data, recording the physical changes of the collections under different environmental conditions, such as the impact of temperature and humidity changes on the material. This data helps museums monitor the condition of collections and the impact of the environment on them in real time, thus providing a scientific basis for the maintenance and protection of the collections. The inventory includes not only exhibited artifacts in the exhibition hall, but also stocked artifacts stored in warehouses, restoration areas, or temporary storage areas. The data collection and inventory process described above also applies to stocked artifacts, so as to achieve unified management and tracking inventory of the museum's entire collection.

[0014] Next, key structural decomposition is performed on the collected contour-related 3D geometric data, regional statistical processing is conducted on the collected surface texture data, and multiple sampling comparisons are performed on the collected material response data to obtain feature sets representing contour, texture, and material respectively. By hierarchically combining these feature sets, a multi-dimensional identity feature vector is generated. This multi-dimensional identity feature vector represents the unique digital identity of the collection, where each dimension describes a specific characteristic of the collection. Then, incremental amplitude verification is performed on the generated multi-dimensional identity feature vector. Once verification is successful, incremental updates of the multi-dimensional identity feature vector are allowed, thus constructing the final dynamic identity fingerprint of the collection. This dynamic identity fingerprint integrates the static features and dynamic change information of the collection, and can be updated in real time with environmental changes, reflecting the status changes of the collection under different environmental conditions. Through this method, the museum can establish a precise, dynamic, and real-time updated collection management system, providing efficient inventory, monitoring, and protection functions to ensure the safety and integrity of the collection.

[0015] In some embodiments, the three-dimensional geometric contour, surface texture, and material response data of the artifact are collected to generate a multi-dimensional identity feature vector, including: The key structure decomposition of the collected three-dimensional geometric contour data is performed to extract the main contour feature set used to characterize the overall morphological stability of the collection; the surface texture data is subjected to regional statistical processing to extract the texture stability feature set that does not depend on local occlusion and illumination changes; multiple sampling comparisons of material response data are performed to extract the material response feature set that remains consistent under repeated acquisition conditions; the main contour feature set, texture stability feature set, and material response feature set are hierarchically combined to form a multi-dimensional identity feature vector with a fixed feature order and structural representation.

[0016] Specifically, the first step involves preprocessing the acquired 3D geometric contour data of the artifacts. This includes noise point removal, outlier correction, and data smoothing of the original point cloud. Rigid registration is then used to unify the 3D geometric contour data into a predefined standard coordinate system. Noise point removal can be performed using a statistical filtering method based on a neighborhood density threshold; outlier correction can be performed using an interpolation reconstruction method based on neighborhood point fitting; and data smoothing can be performed using a moving least squares smoothing method based on local surface fitting. Subsequently, the processed 3D geometric contour data is scaled to eliminate scale differences caused by different acquisition distances or angles. Based on this, key structural decomposition of the 3D geometric contour is performed based on curvature variation analysis and geometric topological relationships. During this process, the curvature values ​​of each sampling point on the contour surface are calculated, including at least one of mean curvature, Gaussian curvature, or principal curvature. At a sampling point, the difference between the curvature value of that sampling point and the mean curvature value of its neighboring sampling points is calculated to obtain the curvature change amplitude. When the curvature change amplitude exceeds a pre-set curvature abrupt change threshold, the corresponding sampling point is marked as a curvature abrupt change point. Adjacent curvature abrupt change points are then spatially clustered, and continuously distributed curvature abrupt change points are aggregated to form curvature abrupt change regions. These regions represent key turning points, structural boundaries, or areas of significant morphological change in the geometric contour of the collection. For these curvature abrupt change regions, contour master features representing the overall shape boundary, main turning surfaces, and key structural proportional relationships of the collection are extracted. These features reflect the stability of the overall shape of the collection. By encoding these contour master features, a contour master feature set can be formed.

[0017] Next, the collected surface texture data of the artifacts underwent preprocessing, including brightness normalization, color space conversion, and noise filtering. The surface texture data was then divided into multiple fixed-scale texture analysis regions according to preset spatial division rules. Brightness normalization was achieved by linearly stretching or histogram equalization of pixel brightness values; color space conversion was performed by converting the original RGB color space to HSV or Lab color space; and noise filtering was achieved using median filtering or Gaussian filtering. Within each texture analysis region, statistical features such as color distribution histogram, texture direction distribution, and frequency energy distribution were calculated. The statistical features of the same region obtained at different acquisition times or under different lighting conditions were compared for consistency. Regions whose features fluctuated beyond a preset threshold across multiple acquisitions were removed, retaining only the statistical results of regions that remained stable under different acquisition conditions. This resulted in a stable texture feature set independent of local occlusion and lighting changes.

[0018] Then, the material response data of the collection was repeatedly sampled under the same acquisition parameters. Each sampling recorded the material response value sequence under temperature, humidity, light, or micro-vibration. After time alignment processing of the material response data obtained from multiple samplings, the response deviation between each sampling sequence was calculated, and response features with response deviations less than a preset stability threshold were marked as consistent response features, thereby extracting a set of material response features that remained consistent under repeated acquisition conditions.

[0019] Finally, the contour master feature set, texture stable feature set, and material response feature set are fused according to a preset hierarchical combination rule. First, the contour master feature set is encoded with high-level features. Then, the texture stable feature set is mapped to the corresponding structural regions and its mid-level features are concatenated. Finally, the material response feature set is appended to the corresponding structural and texture features as low-level features. During the fusion process, fixed feature order identifiers and dimension numbers are assigned to each type of feature, forming a structured, fixed-order feature representation. This generates a multi-dimensional identity feature vector to represent the unique identity of the collection, which is used for subsequent establishment and comparison analysis of the collection's dynamic identity fingerprint.

[0020] In some embodiments, a dynamic identity fingerprint of a collection is established using the multidimensional identity feature vector, including: The multidimensional identity feature vector is used as the initial fingerprint state of the collection's dynamic identity fingerprint. During the inventory operation, if the preset update conditions are met, an incremental update instruction is generated. The inventory data collected during the inventory operation is called according to the incremental update instruction, and the incremental amplitude verification of the initial fingerprint state is performed. If the incremental amplitude verification passes, the incremental update process is allowed.

[0021] Specifically, the generated multidimensional identity feature vector is first stored in a structured manner and marked as the initial fingerprint state of the collection's dynamic identity fingerprint. This initial fingerprint state corresponds to the stable feature set of the collection at the time of its initial archiving or baseline inventory, and serves as the benchmark reference for all subsequent inventory comparisons and updates. During each subsequent inventory operation, the system first determines whether preset update conditions are met. These conditions include, but are not limited to, the inventory time interval reaching a set threshold, the inventory collection quality meeting integrity requirements, the collection's consistency score being within the allowable update range, and no abnormal risk assessment being triggered. When at least one or more preset update conditions are met, the system generates an incremental update instruction to instruct the current collection's dynamic identity fingerprint to undergo candidate update processing. After generating the incremental update instruction, the inventory collection data obtained in this inventory operation is called, and based on the same feature extraction rules as the initial fingerprint state, the corresponding multidimensional identity feature vector is regenerated from the inventory collection data. Subsequently, the newly generated multidimensional identity feature vector is compared and aligned dimension by dimension with the currently stored initial fingerprint state, the change magnitude of each feature dimension is calculated, and a feature change magnitude set is formed. Next, the magnitude of feature changes is compared with a pre-set incremental allowable threshold to determine whether geometric contour features, texture stability features, and material response features are within the allowable range of change, thus performing incremental magnitude verification of the initial fingerprint state. When the magnitude of change in a certain feature dimension exceeds the corresponding incremental allowable threshold, the incremental magnitude verification is deemed successful. At this point, the system allows incremental update processing, writing the verified feature changes from the data collected in this inventory into the current dynamic identity fingerprint of the collection incrementally, updating the value of the corresponding feature dimension, and simultaneously recording the update time, update source, and update magnitude information. Conversely, if the incremental magnitude verification fails, the update operation is rejected, and the original dynamic identity fingerprint state remains unchanged. Through the above process, the adaptive evolution of the collection's identity fingerprint is achieved, maintaining the long-term stability of identity recognition while being compatible with reasonable changes brought about by the natural aging of cultural relics and compliance interventions, thus improving the robustness and practicality of intelligent inventory.

[0022] The exhibition space is constructed by taking the collections, display stands and fixed components in the exhibition space as nodes, and establishing edges with the relative position constraints, accessibility constraints and synchronous change constraints between nodes.

[0023] Specifically, the system first measures and models the exhibition space's structure, display cases, stands, and fixed components to establish a unified exhibition space coordinate system. Each artifact, its corresponding stand, and fixed components such as walls, display cases, and support frames are abstracted as independent spatial nodes, each assigned a unique node identifier and spatial coordinate information. Then, based on the spatial coordinate information, the relative spatial relationship between any two nodes is calculated, including distance, azimuth, height difference, and orientation. This relative spatial relationship is compared with preset installation specifications. When the relative spatial relationship meets the constraints in the installation specifications, relative position constraint edges are established between the corresponding nodes to characterize the stable spatial positional relationship that the nodes should maintain. After establishing the relative position constraint edges, the system determines the accessibility constraints between nodes based on the physical structure and management rules of the exhibition space. Specifically, it considers the display case's closed state, passageway layout, protective structure, and management access information to determine whether the artifact node can directly contact or exchange positions with other nodes without external intervention. When there is no reachable path between two nodes under normal exhibition conditions, reachability constraint edges are established between the nodes to characterize the restriction that they cannot be directly interchanged or moved in physical space. Simultaneously, for node groups with interconnected relationships, synchronous change constraints are constructed. Specifically, this includes relationships between collection nodes and exhibition platform nodes located on the same display stand, and between multiple nodes installed on the same fixed component. When any node changes position, its associated nodes should exhibit synchronous changes in spatial position or orientation. Based on this, synchronous change constraint edges are established between related nodes. After the above constraint edges are established, all nodes and their corresponding relative position constraint edges, reachability constraint edges, and synchronous change constraint edges are organized into a graph structure data model, forming an exhibition space constraint graph. Nodes describe entities in the exhibition space, and edges describe the spatial constraint relationships that should be maintained between nodes over a long period. This exhibition space constraint graph serves as the foundational structure for subsequent inventory analysis and spatial consistency determination, used to detect anomalies in the spatial position, reachability, and interconnected changes of collections.

[0024] Perform microenvironmental data collection on the collection and establish microenvironmental evolution data including temperature, humidity, light, and vibration data.

[0025] Specifically, firstly, microenvironment monitoring units are deployed at the exhibition or storage locations of the artifacts. These units include, but are not limited to, temperature, humidity, light, and vibration sensors. The installation locations and sampling accuracy of the sensors are then determined based on the type of artifact and conservation requirements. Subsequently, all sensors undergo unified calibration and time synchronization to ensure that different types of microenvironment data are collected under the same time reference. During the operation of the intelligent inventory system, temperature, humidity, light intensity, and vibration amplitude data around the artifacts are continuously collected according to a preset sampling cycle, and each data point is appended with a corresponding timestamp and artifact identification information. After real-time data collection, the temperature, humidity, light, and vibration data are preprocessed, including outlier removal, missing data completion, and data smoothing, to eliminate the impact of transient interference or sensor jitter. Then, the processed microenvironment data is sorted and segmented chronologically, and various microenvironment parameters are aggregated and calculated based on continuous time windows to form time-series data reflecting the changing trends of the microenvironment. Based on this, the time series of temperature, humidity, light, and vibration of the same collection in different time periods are correlated and integrated to construct a multidimensional microenvironment state vector. This vector is then continuously spliced ​​together in chronological order to form microenvironment evolution data that describes the changes in the environment surrounding the collection over time. This microenvironment evolution data not only records the instantaneous values ​​of various environmental parameters but also includes derived characteristics such as the magnitude, rate of change, and duration of their changes. It is used to characterize the long-term stability and short-term fluctuations of the microenvironment in which the collection is located, providing basic data support for subsequent collection consistency analysis, environmental driving model construction, and anomaly risk assessment.

[0026] By jointly analyzing the dynamic identity fingerprints of the collections, the exhibition space constraint diagrams, and the microenvironment evolution data, the residuals with historical data and related collections are calculated to form a collection consistency score.

[0027] Specifically, the system first retrieves the current dynamic identity fingerprint of the artifact to be inventoried, the pre-constructed exhibition space constraint map, and the microenvironment evolution data of the artifact. Simultaneously, it acquires historical inventory data and data of related artifacts with spatial or environmental connections, serving as a reference benchmark for joint analysis. Then, using the artifact's dynamic identity fingerprint as the core, the system compares the current fingerprint with historical fingerprints dimension-by-dimensionally to form a first consistency score reflecting the degree of change in the artifact's identity. Based on this, the spatial consistency of the artifact is analyzed using the exhibition space constraint map, calculating deviation values ​​corresponding to various spatial constraints to form a second consistency score, which characterizes the degree of anomaly in the artifact's spatial location and interrelationships. Simultaneously, the system performs time-series analysis on the microenvironment evolution data, aligning the microenvironment change trajectory within the current inventory period with the historical microenvironment evolution trajectory, calculating multiple deviation values, and forming a third consistency score. Subsequently, the obtained first consistency score, second consistency score, and third consistency score are mapped to the same evaluation space to calculate the collection consistency score. The higher the collection consistency score, the closer the collection status is to the historical stable state and the state of the associated environment. Conversely, the lower the collection consistency score, the higher the abnormal risk of the collection. This serves as the core basis for triggering subsequent intelligent inventory strategies and judging abnormal risks.

[0028] In some embodiments, the dynamic identity fingerprint of the collection, the exhibition space constraint diagram, and the microenvironment evolution data are jointly analyzed to calculate the residuals with historical data and related collections, forming a collection consistency score, including: After accessing the historical dynamic identity fingerprint of the collection, deviation residuals are calculated based on the collection's dynamic identity fingerprints to establish a first consistency score. Based on the exhibition space constraint diagram, the relative positional relationships between the collection and adjacent collections, display stands, and fixed components are calculated. The consistency of these relative positional relationships is calculated using preset spatial conditions to establish spatial consistency residuals. A second consistency score is established based on these spatial consistency residuals. Based on the microenvironment evolution data, the trajectory fitting deviation between the microenvironment change trajectory corresponding to the collection and the historical microenvironment evolution trajectory is calculated to establish a first deviation. The associated microenvironment change trajectories of related collections are obtained. Similarity analysis is performed on the associated microenvironment change trajectories and the microenvironment change trajectories to establish a second deviation. A third consistency score is constructed based on the first and second deviations. The first, second, and third consistency scores are mapped to the same evaluation space to establish a collection consistency score.

[0029] Specifically, the process begins by retrieving the historical dynamic identity fingerprint data corresponding to the target collection. The dynamic identity fingerprint of the collection generated during the current inventory cycle is then aligned dimension-by-dimensionally with the retrieved historical dynamic identity fingerprint. During this process, for each feature dimension in the multi-dimensional identity feature vector, the difference between the current feature value and the historical feature value is calculated, and the calculated difference is normalized to form a set of identity feature deviation residuals. This set of identity feature deviation residuals is then weighted and summarized to obtain the first deviation residual, which characterizes the stability of the collection's identity. Based on a preset residual-score mapping function, the first deviation residual is converted into a first consistency score, reflecting the degree of consistency of the collection's identity features relative to its historical state. This residual-score mapping function uses a non-linear mapping form, such as exp(−R / T), where R is the first deviation residual and T is the residual sensitivity coefficient, used to adjust the response speed of the score to changes in the residual.

[0030] Based on this, the spatial consistency of the collection is calculated using the exhibition space constraint diagram. In this process, the system extracts adjacent collection nodes, display stand nodes, and fixed component nodes directly associated with the target collection node from the exhibition space constraint diagram, and reads their corresponding relative position constraints, accessibility constraints, and synchronous change constraints. Subsequently, based on the spatial positioning data obtained from the current inventory, the actual relative positional relationship between the target collection and each associated node is calculated, including distance, direction, height difference, and posture relationship. The actual relative positional relationship is then compared item by item with the spatial conditions recorded in the constraint diagram. When the actual relative positional relationship deviates from the spatial conditions, the corresponding spatial deviation value is calculated and recorded as a spatial consistency residual. Afterwards, all spatial consistency residuals are weighted and the weighted result is mapped to a second consistency score, used to characterize the stability of the collection in terms of spatial position and linkage relationships.

[0031] Simultaneously, consistency analysis is performed on the microenvironment evolution data. This process first extracts the microenvironmental change trajectory of the target artifact within the current inventory period and aligns it with the artifact's historical microenvironmental evolution trajectory using time alignment and trajectory fitting. By calculating the fitting error, trend deviation, and fluctuation amplitude difference, a first deviation reflecting the degree of difference between the current and historical environmental changes is obtained. The fitting error is quantified by using the mean square error to compare the overall numerical values ​​of the current and historical trajectories. The trend deviation is quantified by calculating the absolute difference between the slope of the linearly fitted trend of the current trajectory and the slope of the trend of the historical trajectory. The fluctuation amplitude difference is quantified by calculating the absolute difference between the standard deviation of the current trajectory and the standard deviation of the historical trajectory. Subsequently, the microenvironmental change trajectories of related artifacts located in the same exhibition area as the target artifact are obtained, and a similarity analysis is performed between these related microenvironmental change trajectories and the target artifact's microenvironmental change trajectory. The Euclidean distance between the two is calculated, forming a second deviation. Finally, a third consistency score is mapped based on the weighted results of the first and second deviations, reflecting the consistency level of the target artifact's microenvironmental changes relative to its own historical state and the overall environmental state of the region.

[0032] Finally, the system maps the first consistency score, the second consistency score, and the third consistency score to the same evaluation space, normalizes the scale and adjusts the weights of each score, and integrates them according to the preset fusion rules to obtain a collection consistency score that comprehensively represents the consistency of the collection's identity characteristics, spatial status, and environmental status. This collection consistency score serves as the core evaluation indicator for triggering subsequent intelligent inventory strategies and judging abnormal risks. It can not only help museums monitor the status of collections in real time, but also trigger risk warnings in a timely manner when anomalies occur, ensuring the safety and integrity of the collections.

[0033] In some embodiments, the dynamic identity fingerprint and micro-environment evolution data of the collection are used to determine their consistency, and a collaborative constraint term for the consistency score of the collection is established. The collaborative consistency determination includes: Based on the microenvironment evolution data, a collection environment change-driven model is constructed. This model describes the influence of temperature, humidity, light, and vibration on the material response characteristics of the collection. According to the collection environment change-driven model, the allowable variation range of the material response characteristics in the dynamic identity fingerprint of the collection is adaptively predicted to establish an identity change prediction interval. The currently collected dynamic identity fingerprint of the collection is matched with the identity change prediction interval to establish a collaborative anomaly residual. The collaborative anomaly residual is introduced as a high-priority constraint into the calculation process of the collection consistency score to execute consistency calculation constraints.

[0034] Specifically, firstly, based on the collected and organized microenvironment evolution data, the change sequences of environmental parameters such as temperature, humidity, light intensity, and vibration amplitude are extracted in chronological order, and simultaneously, the change sequences of material response features in the dynamic identity fingerprint of the collection within the corresponding time period are extracted. Subsequently, the environmental parameter change sequences and material response feature change sequences are time-aligned, and a mapping relationship model between environmental parameters and material response features is established through regression analysis, correlation analysis, or causal modeling.

[0035] Taking multivariate linear regression as an example, the microenvironment evolution data is first divided into time windows using a fixed time interval Δt, such as 5 minutes or 10 minutes. Within each time window, the changes in temperature, humidity, light intensity, and vibration amplitude are calculated. This change is defined as the mean of the current time window minus the mean of the previous time window, thus obtaining the temperature change ΔT, humidity change ΔH, light intensity change ΔL, and vibration change ΔV. Simultaneously, within the same time window, the change ΔM of the material response feature corresponding to the dynamic identity fingerprint of the artifact is extracted. This material response feature can be a quantified numerical indicator such as reflectivity change, dielectric response change, or material micro-vibration response change. After obtaining ΔT, ΔH, ΔL, ΔV, and ΔM within continuous time windows, a clear functional relationship is constructed: ΔM = a·ΔT + b·ΔH + c·ΔL + d·ΔV + ε, where a, b, c, and d are the environmental influence coefficients to be determined, and ε is the residual term. Subsequently, data from multiple consecutive time windows are substituted into the aforementioned functional relationship, and the least squares method is used to calculate the environmental impact coefficients a, b, c, and d, minimizing the sum of the squared errors between the predicted values ​​and the actual ΔM across all time windows. Through this calculation process, the quantitative impact coefficient of each environmental parameter on the material response characteristics is directly obtained, thus clarifying the strength and direction of the influence of different environmental factors on the material response.

[0036] After obtaining the environmental impact coefficients, the model's effectiveness is validated. Specifically, the calculated values ​​of a, b, c, and d are re-substituted into historical data, and the predicted material response change value ΔM′ is calculated for each time window. This value is then compared window-by-window with the actual collected ΔM, and the absolute value distribution of the prediction error is statistically analyzed. When the proportion of prediction errors within a preset allowable range reaches a threshold requirement, the stability of the regression relationship is confirmed. Subsequently, based on historical micro-environmental data, the maximum variation range of each environmental parameter under normal exhibition conditions is statistically analyzed. This range is then substituted into the regression model to calculate the theoretical maximum and minimum changes in material response characteristics under different combinations of environmental changes, thereby clarifying the response boundaries of material response characteristics under current environmental conditions. Through the aforementioned steps of clearly defined time window division, change calculation, linear function construction, parameter solving, and error verification, a collection environmental change-driven model is ultimately formed. This model describes, in the form of explicit mathematical relationships, the degree, direction, and allowable variation range of the influence of temperature, humidity, light, and vibration changes on the material response characteristics of the collection, providing a directly executable computational basis for the subsequent construction of identity change prediction ranges and collaborative anomaly determination.

[0037] After constructing the environmental change-driven model for the collections, the system uses micro-environmental evolution data within the current inventory cycle as model input during each inventory analysis. Based on the environmental change-driven model, it predicts the theoretical change range of each material response feature in the dynamic identity fingerprint of the collections under the current environmental conditions. By performing interval expansion processing on the prediction results, an identity change prediction interval containing upper and lower limits is generated to characterize the reasonable range of allowable changes in material response features under the current micro-environmental change conditions. Then, each material response feature in the currently collected dynamic identity fingerprint of the collection is matched and judged with the corresponding identity change prediction interval. When the actual change value of a certain material response feature exceeds the upper or lower limit of the prediction interval, its excess magnitude is calculated and recorded as a co-anomaly residual. Then, the excess magnitudes corresponding to all material response features are summarized to form an overall co-anomaly residual index, which is used to reflect the degree of inconsistency between the current material response change of the collections and the driving relationship of environmental change. Finally, the co-abnormal residuals are introduced as high-priority constraints into the calculation of the collection consistency score. When performing consistency score fusion calculation, the co-abnormal residuals are given a higher weight than other residual items. When the co-abnormal residuals exceed the preset threshold, the collection consistency score is directly reduced. This ensures that even if other consistency score dimensions are within the normal range, abnormal changes can still be effectively amplified and identified, thereby ensuring that abnormalities in the collection materials can be accurately identified and included in the risk assessment system at an early stage.

[0038] The inventory strategy is triggered based on the consistency score of the collection, and an intelligent inventory operation is performed.

[0039] Specifically, after calculating the consistency score of the collections, the consistency scores obtained within the current inventory period are first summarized with previous consistency scores to form a time evolution sequence of the consistency scores. Then, based on this time evolution sequence, the consistency scores are compared item by item with stability judgment rules to determine whether they meet the conditions for continuous stability, short-term fluctuations, or abnormal evolution. When these conditions are not met, the system determines that the current state is abnormal. In this case, an intelligent inventory strategy is configured according to this abnormal evolution state, and the intelligent inventory operation is executed accordingly. After the inventory strategy is triggered, the corresponding acquisition devices, analysis modules, and data processing flows are automatically invoked according to the corresponding strategy configuration to complete the intelligent inventory operation. The inventory execution process, acquisition results, and the reason for strategy triggering are recorded, providing data support for subsequent anomaly risk analysis, inventory strategy optimization, and management decisions.

[0040] In some embodiments, an inventory strategy is triggered based on the consistency score of the collection to perform an intelligent inventory operation, including: The consistency scores of the collections are recorded over multiple inventory cycles to construct a temporal evolution sequence of the consistency scores. Based on the temporal evolution sequence of the consistency scores, the rate of change, fluctuation amplitude, and recovery characteristics of the scores are extracted to establish an evolutionary feature set. The evolutionary feature set is compared item by item with preset stability judgment rules to determine whether the consistency scores of the collections meet the conditions of continuous stability, short-term fluctuation, or abnormal evolution. When the rate of change of the scores exceeds a preset change threshold, or when the score fluctuation does not show a regression trend within a preset inventory cycle, it is determined to be an abnormal evolution state. After configuring an intelligent inventory strategy according to the abnormal evolution state, the intelligent inventory operation is executed.

[0041] Specifically, after each inventory check is completed and a consistency score is generated, the consistency score is bound and stored with the corresponding inventory timestamp in chronological order, and continuously recorded across multiple consecutive inventory cycles. This forms a time-indexed evolution sequence of the consistency score, which comprehensively records the changes in the consistency status of the collection across different inventory cycles. This serves as the foundational data source for subsequent trend analysis, enabling continuous perception and intelligent tracking of changes in the collection's status. After constructing the consistency score time evolution sequence, the system performs feature extraction processing based on it. During this process, the rate of change of the score is calculated using the score difference between adjacent inventory cycles to characterize how quickly the consistency score changes over time. The difference between the maximum and minimum scores within a preset time window is then used to calculate the score fluctuation amplitude, reflecting the stability level of the consistency score. Simultaneously, the system analyzes the recovery of the score after a decline, extracting recovery features, including recovery duration, recovery slope, and whether it returns to a historical stable range. This forms an evolutionary feature set containing the rate of change, fluctuation amplitude, and recovery characteristics. Subsequently, the evolutionary feature set is compared item by item with pre-defined stability judgment rules. These stability judgment rules include at least threshold conditions for continuous stability, short-term fluctuations, and abnormal evolution. When the rate of change of the score is consistently below the change threshold and the fluctuation amplitude is within the allowable range, the consistency score is determined to meet the continuous stability condition. When the score fluctuates in a short period of time but returns to the stable range within the preset inventory period, it is determined to meet the short-term fluctuation condition. When the rate of change of the score exceeds the preset change threshold, or the score fluctuation does not show a return trend within the preset inventory period, or even shows a continuous downward trend, the consistency score of the collection is determined to be in an abnormal evolution state, used to characterize abnormal change nodes or abnormal evolution stages that occur in the collection during intelligent tracking. After being determined to be in an abnormal evolution state, the system automatically configures the corresponding intelligent inventory strategy according to the type and severity of the abnormal evolution, specifically including increasing the inventory frequency, increasing the collection dimensions, improving the collection accuracy, or triggering a special inventory process. Subsequently, the system automatically calls the corresponding inventory equipment and analysis modules according to the configured intelligent inventory strategy, performs intelligent inventory operations, and records the execution results in association with the abnormal evolution judgment results. In this way, based on the continuous intelligent tracking of changes in the status of the collection, it provides data support for subsequent abnormal risk assessment, inventory strategy optimization and management decision-making.

[0042] The system uses intelligent inventory results and consistency scores to report abnormal risks in the collection.

[0043] Specifically, after the intelligent inventory operation is completed, the results obtained during the operation are first collected, including the latest 3D geometric contour data, surface texture data, material response data, spatial position verification results, and microenvironment verification results of the collections. Simultaneously, the corresponding collection consistency score and its source information are obtained. Then, the intelligent inventory operation results and the collection consistency score are jointly verified to determine their consistency in terms of anomaly type, direction, and severity, thus forming a joint verification result. Based on the joint verification result, preset anomaly risk judgment rules are invoked to perform a graded analysis, classifying the anomaly status of the collections into different anomaly risk levels. This anomaly risk level includes at least low risk, suspected risk, and high risk. When the consistency score is within the attention range and the inventory operation results show only slight deviations, it is judged as a low-risk state; when the consistency score continues to decline and structural or environmental anomalies appear in the inventory operation results, it is judged as a suspected risk state; when the consistency score is below the anomaly threshold and significant identity, spatial, or material anomalies are detected in the inventory operation results, it is judged as a high-risk state. After determining the level of abnormal risk, the corresponding abnormal risk information of the collection is generated. This abnormal risk information is then associated with the collection identifier, consistency score value, abnormal type, risk level, and generation timestamp to form a standardized abnormal risk output result. This abnormal risk output result can be used to trigger alarm prompts, risk reporting, or subsequent manual verification processes, thereby realizing the reporting of abnormal risks of collections based on the results of intelligent inventory operation and the consistency score of the collection.

[0044] In some embodiments, the risk of anomalies in collections is reported using the results of intelligent inventory operations and collection consistency scores, including: The abnormal risk recording unit is activated, and the results of the intelligent inventory operation, the consistency score of the collection, the abnormal risk judgment result, and the risk generation timestamp are combined to establish an abnormal risk evidence data package; the abnormal risk evidence data package is subjected to structured summary processing to generate risk fingerprint data for characterizing the abnormal risk formation process; the risk fingerprint data is written into the block record structure and linked with the previous abnormal risk record to form an abnormal risk evolution block.

[0045] Specifically, after the system completes the assessment of anomaly risks in the collections, it first activates the anomaly risk recording unit to uniformly collect and process relevant data generated during the current inventory period. In this process, the anomaly detection information, corresponding collection consistency score values, anomaly risk assessment results, and risk generation timestamps involved in the current intelligent inventory operation are combined and encapsulated, and a unique data identifier is assigned to this combined data, thereby constructing a complete anomaly risk evidence data package. This anomaly risk evidence data package is used to objectively record the original basis and assessment results when the anomaly risk occurred. After the anomaly risk evidence data package is constructed, it undergoes structured summary processing, specifically including extracting key information such as the type of inventory anomaly, the amount of change in anomaly characteristics, the change in consistency score, and the assessment conclusion, and encoding and compressing it according to a preset field structure. Based on this, risk fingerprint data that uniquely represents the formation process of this anomaly risk is generated through hash calculation or feature mapping. This risk fingerprint data is used for integrity verification and rapid indexing of the anomaly risk evidence data package. The system then writes the generated risk fingerprint data into a block record structure. During the writing process, it appends the block identifier of the preceding abnormal risk record to the risk fingerprint data and links the current risk fingerprint data with the preceding risk fingerprint data through a chain reference, thereby forming a continuous abnormal risk evolution block. Through the above chain association structure, the system can achieve tamper-proof storage of abnormal risk records and traceability of the evolution process, providing reliable data support for subsequent risk source tracing analysis, responsibility determination, and management decisions.

[0046] In some embodiments, the risk of anomalies in collections is reported using the results of intelligent inventory operations and collection consistency scores, including: The results of the intelligent inventory operation and the consistency score of the collection are jointly verified, and the joint verification result is output. Based on the joint verification result, the abnormal warning level is matched, and the abnormal risk of the collection is reported using the abnormal warning level matching result.

[0047] Specifically, after completing the intelligent inventory operation and obtaining the corresponding collection consistency score, the various detection conclusions identified in the intelligent inventory operation results are first structured and organized, including identity feature comparison results, spatial location verification results, micro-environment review results, and information on whether there are any exceeding limits. These detection conclusions are then correlated with the collection consistency scores generated within the same inventory cycle. Subsequently, the anomalies identified in the intelligent inventory operation results are classified according to their source, including at least identity anomalies, spatial anomalies, and environmental anomalies. For each type of anomaly, corresponding anomaly type labels, anomaly direction identifiers, and anomaly intensity values ​​are extracted. The anomaly direction indicates whether the anomaly deviates from historical state, spatial constraints, or environmentally driven expectations; the anomaly intensity value quantifies the extent to which the anomaly exceeds the allowable threshold; identity anomalies characterize inexplicable differences between the collection's three-dimensional geometric contour, surface texture, or material response features and its historical dynamic identity fingerprint, used to identify the risk of the collection being swapped, counterfeited, or replaced; spatial anomalies characterize changes in the position or orientation of the collection when accessibility constraints or synchronous change constraints are not met, used to identify the risk of the collection being illegally moved or stolen.

[0048] Simultaneously, the consistency score of the collection is broken down into corresponding sub-scores, including a first consistency score, a second consistency score, and a third consistency score. The main deviation type and overall deviation intensity reflected by the consistency score are determined based on the decrease in each sub-score. Then, the system performs consistency judgment processing based on anomaly type matching. Specifically, when an identity anomaly is found in the intelligent inventory operation results, the system checks whether the first consistency score decreases synchronously; when a spatial anomaly is found, it checks whether the second consistency score decreases synchronously; and when an environmental or material anomaly is found, it checks whether the third consistency score decreases synchronously, thus completing the anomaly type-level matching judgment. Through this anomaly type matching process, it is possible to verify whether the detected anomalies are mutually corroborated at both the identity feature level and the consistency score level, thereby improving the reliability of identifying abnormal behavior in collections and avoiding misjudgments caused by a single detection result.

[0049] After completing the anomaly type matching, the trend direction of each sub-score in the consistency score is compared with the anomaly direction indicator of the inventory anomaly item. For example, whether it simultaneously exhibits continuous deviation, abrupt deviation, or gradual deviation. When the directions are consistent, it is marked as a successful direction match; otherwise, it is marked as a mismatch. Next, anomaly intensity consistency judgment is performed. The anomaly intensity value of each anomaly item in the inventory operation results is compared with the decrease range of the consistency score through interval mapping. When the anomaly intensity is within the deviation range corresponding to the consistency score, it is determined that the intensity match is successful. Through the dual consistency judgment of anomaly direction and anomaly intensity, normal changes caused by natural environmental fluctuations and abnormal changes caused by human intervention can be further distinguished, enhancing the ability to identify theft, substitution, and counterfeiting from a technical perspective.

[0050] After completing anomaly type matching, anomaly direction matching, and anomaly intensity matching, the three types of matching results are quantified and weighted according to preset joint verification scoring rules to generate a joint verification credibility index. When the number of matching items reaches a preset proportion or the joint verification credibility index exceeds a preset threshold, the joint verification is deemed successful, and a high consistency joint verification result is output. When there are insufficient matching items or obvious inconsistencies, a low consistency or conflicting joint verification result is output. This joint verification result is used to quantitatively characterize the credibility and consistency of the current collection anomaly judgment and serves as the direct basis for subsequent anomaly warning level matching and anomaly risk reporting.

[0051] After obtaining the joint verification results, the system invokes the anomaly warning level matching rules to map the joint verification results to the corresponding anomaly warning levels. Finally, based on the anomaly warning level matching results, it generates corresponding anomaly risk output information for the collection and reports the anomaly risk level, anomaly source, and relevant judgment criteria in conjunction with the results. This enables intelligent identification and early warning of risks such as theft, illegal movement, substitution, and counterfeiting of collections without relying on manual intervention, achieving anomaly risk reporting based on joint verification results.

[0052] Furthermore, in the intelligent tracking and inventory method for museum collections, the system continuously records and associates inventory data generated by each collection during various inventory cycles based on a dynamic identity fingerprint established for each item, forming a traceable record of the collection's status evolution. Specifically, after each inventory operation, the system binds and stores the dynamic identity fingerprint, consistency score, spatial location verification result, microenvironment evolution data, and anomaly risk judgment result collected within the corresponding inventory cycle, along with the inventory timestamp and the collection's unique identifier, forming structured inventory record nodes. Subsequently, the system can sequentially associate multiple inventory record nodes according to time order, constructing a collection tracing record chain with time as the primary index, used to describe the changes in the collection's identity status, spatial status, and environmental status at different time points. When an anomaly risk event occurs, the anomaly risk evidence data package and the corresponding risk fingerprint can be written into the tracing record chain, marking the anomaly occurrence time, anomaly type, and associated inventory cycle, thereby forming a clear anomaly evolution node in the tracing record.

[0053] During intelligent tracing, the system retrieves the corresponding inventory record nodes from the tracing record chain based on the user-input item identifier, time range, or abnormal event identifier. It then sequentially reconstructs the changes in the item's identity characteristics, spatial location trajectory, and micro-environment within the specified time range. By comparing and analyzing adjacent inventory record nodes, the system can trace the start time, path, and key influencing factors of the item's status changes, enabling intelligent tracing of the item's "whereabouts," "status change process," and the causes of any anomalies.

[0054] Through the aforementioned recording and tracing mechanisms, each collection item possesses a continuous and traceable digital identity and status trajectory, thereby establishing a unique digital identity file for the collection. This supports accurate querying and tracing of the collection's historical status and abnormal evolution process during post-event verification, responsibility determination, and management auditing.

[0055] In summary, the intelligent tracking and inventory method for museum collections provided by this invention has the following technical effects: The process involves collecting 3D geometric contours, surface textures, and material response data of artifacts to generate multidimensional identity feature vectors. These vectors are then used to establish dynamic identity fingerprints for the artifacts. Artifacts, display stands, and fixed components within the exhibition space are treated as nodes, and edges are established using relative position constraints, accessibility constraints, and synchronous change constraints between nodes to construct an exhibition space constraint graph. Microenvironmental data collection of the artifacts is performed, establishing microenvironmental evolution data including temperature, humidity, light, and vibration data. The dynamic identity fingerprints, exhibition space constraint graph, and microenvironmental evolution data are jointly analyzed to calculate residuals with historical data and related artifacts, forming an artifact consistency score. Based on the artifact consistency score, an inventory strategy is triggered, and intelligent inventory operations are performed. The results of the intelligent inventory operation and the artifact consistency score are used to report artifact anomaly risks. This achieves the technical effect of intelligent assessment of artifact consistency through multidimensional feature fusion, spatial constraint modeling, and microenvironmental evolution analysis, improving inventory efficiency and accuracy, and strengthening the intelligent management of the entire artifact protection process.

[0056] Example 2, as Figure 2 This is a schematic diagram of the intelligent tracking and inventory system for museum collections of the present invention. For example, Figure 1 The flowchart of the intelligent tracking and inventory method for museum collections of this invention can be seen as follows: Figure 2 The structure shown is implemented.

[0057] Based on the same concept as the intelligent tracking and inventory method for museum collections described in the embodiments, the present invention also provides an intelligent tracking and inventory system for museum collections, comprising: Fingerprint Establishment Module 11: Collects the three-dimensional geometric contour, surface texture, and material response data of the collection, generates a multi-dimensional identity feature vector, and uses the multi-dimensional identity feature vector to establish a dynamic identity fingerprint of the collection; Constraint Construction Module 12: Uses the collection, display stand, and fixed components in the exhibition space as nodes, and establishes edges with relative position constraints, accessibility constraints, and synchronous change constraints between nodes to construct an exhibition space constraint graph; Data Acquisition Module 13: Performs micro-environment data acquisition of the collection, and establishes micro-environment evolution data including temperature, humidity, light, and vibration data; Joint Analysis Module 14: Performs joint analysis of the dynamic identity fingerprint of the collection, the exhibition space constraint graph, and the micro-environment evolution data, calculates the residual with historical data and related collections, and forms a collection consistency score; Intelligent Inventory Module 15: Triggers an inventory strategy based on the collection consistency score and executes an intelligent inventory operation; Risk Reporting Module 16: Reports abnormal risks of the collection using the results of the intelligent inventory operation and the collection consistency score.

[0058] In some embodiments, the fingerprint establishment module 11 includes: The key structure decomposition of the collected three-dimensional geometric contour data is performed to extract the main contour feature set used to characterize the overall morphological stability of the collection; the surface texture data is subjected to regional statistical processing to extract the texture stability feature set that does not depend on local occlusion and illumination changes; multiple sampling comparisons of material response data are performed to extract the material response feature set that remains consistent under repeated acquisition conditions; the main contour feature set, texture stability feature set, and material response feature set are hierarchically combined to form a multi-dimensional identity feature vector with a fixed feature order and structural representation.

[0059] In some embodiments, the fingerprint establishment module 11 includes: The multidimensional identity feature vector is used as the initial fingerprint state of the collection's dynamic identity fingerprint. During the inventory operation, if the preset update conditions are met, an incremental update instruction is generated. The inventory data collected during the inventory operation is called according to the incremental update instruction, and the incremental amplitude verification of the initial fingerprint state is performed. If the incremental amplitude verification passes, the incremental update process is allowed.

[0060] In some embodiments, the joint analysis module 14 includes: After accessing the historical dynamic identity fingerprint of the collection, deviation residuals are calculated based on the collection's dynamic identity fingerprints to establish a first consistency score. Based on the exhibition space constraint diagram, the relative positional relationships between the collection and adjacent collections, display stands, and fixed components are calculated. The consistency of these relative positional relationships is calculated using preset spatial conditions to establish spatial consistency residuals. A second consistency score is established based on these spatial consistency residuals. Based on the microenvironment evolution data, the trajectory fitting deviation between the microenvironment change trajectory corresponding to the collection and the historical microenvironment evolution trajectory is calculated to establish a first deviation. The associated microenvironment change trajectories of related collections are obtained. Similarity analysis is performed on the associated microenvironment change trajectories and the microenvironment change trajectories to establish a second deviation. A third consistency score is constructed based on the first and second deviations. The first, second, and third consistency scores are mapped to the same evaluation space to establish a collection consistency score.

[0061] In some embodiments, the joint analysis module 14 includes: Based on the microenvironment evolution data, a collection environment change-driven model is constructed. This model describes the influence of temperature, humidity, light, and vibration on the material response characteristics of the collection. According to the collection environment change-driven model, the allowable variation range of the material response characteristics in the dynamic identity fingerprint of the collection is adaptively predicted to establish an identity change prediction interval. The currently collected dynamic identity fingerprint of the collection is matched with the identity change prediction interval to establish a collaborative anomaly residual. The collaborative anomaly residual is introduced as a high-priority constraint into the calculation process of the collection consistency score to execute consistency calculation constraints.

[0062] In some embodiments, the intelligent inventory module 15 includes: The consistency scores of the collections are recorded over multiple inventory cycles to construct a temporal evolution sequence of the consistency scores. Based on the temporal evolution sequence of the consistency scores, the rate of change, fluctuation amplitude, and recovery characteristics of the scores are extracted to establish an evolutionary feature set. The evolutionary feature set is compared item by item with preset stability judgment rules to determine whether the consistency scores of the collections meet the conditions of continuous stability, short-term fluctuation, or abnormal evolution. When the rate of change of the scores exceeds a preset change threshold, or when the score fluctuation does not show a regression trend within a preset inventory cycle, it is determined to be an abnormal evolution state. After configuring an intelligent inventory strategy according to the abnormal evolution state, the intelligent inventory operation is executed.

[0063] In some embodiments, the risk reporting module 16 includes: The abnormal risk recording unit is activated, and the results of the intelligent inventory operation, the consistency score of the collection, the abnormal risk judgment result, and the risk generation timestamp are combined to establish an abnormal risk evidence data package; the abnormal risk evidence data package is subjected to structured summary processing to generate risk fingerprint data for characterizing the abnormal risk formation process; the risk fingerprint data is written into the block record structure and linked with the previous abnormal risk record to form an abnormal risk evolution block.

[0064] In some embodiments, the risk reporting module 16 includes: The results of the intelligent inventory operation and the consistency score of the collection are jointly verified, and the joint verification result is output. Based on the joint verification result, the abnormal warning level is matched, and the abnormal risk of the collection is reported using the abnormal warning level matching result.

[0065] In embodiment three, the present invention also provides a computer-readable storage medium that can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the intelligent tracking and inventory method for museum collections in the embodiments of the present invention, thereby realizing the above-mentioned intelligent tracking and inventory method for museum collections.

[0066] It should be understood that the embodiments disclosed in this invention and the above description enable those skilled in the art to implement this invention. However, this invention is not limited to the embodiments mentioned above. It should be understood that those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this invention, and should all be included within the protection scope of this invention.

Claims

1. A method for intelligent tracking and inventorying museum collections, characterized in that: The method includes: Collect the three-dimensional geometric contour, surface texture and material response data of the collection, generate a multi-dimensional identity feature vector, and use the multi-dimensional identity feature vector to establish a dynamic identity fingerprint of the collection. The exhibition space constraint diagram is constructed by taking the collections, exhibition stands and fixed components in the exhibition space as nodes, and establishing edges with the relative position constraints, accessibility constraints and synchronous change constraints between nodes. Perform microenvironmental data collection on the collection and establish microenvironmental evolution data including temperature, humidity, light, and vibration data; The dynamic identity fingerprint of the collection, the exhibition space constraint diagram, and the micro-environment evolution data are jointly analyzed to calculate the residuals with historical data and related collections, and a collection consistency score is formed. Based on the consistency score of the collection, an inventory strategy is triggered to execute intelligent inventory operations; The system uses intelligent inventory results and consistency scores to report abnormal risks in the collection.

2. The intelligent tracking and inventory method for museum collections as described in claim 1, characterized in that, The dynamic identity fingerprints of the collections, the exhibition space constraint diagrams, and the microenvironment evolution data are jointly analyzed to calculate the residuals with historical data and related collections, forming a collection consistency score, including: After retrieving the historical dynamic identity fingerprint of the collection, the deviation residual is calculated based on the dynamic identity fingerprint of the collection to establish a first consistency score; Based on the exhibition space constraint diagram, the relative positional relationship between the collection and adjacent collections, display stands and fixed components is calculated. The consistency of the relative positional relationship is calculated using preset spatial conditions to establish spatial consistency residuals. A second consistency score is established based on the spatial consistency residuals. Based on the microenvironment evolution data, the trajectory fitting deviation between the microenvironment change trajectory corresponding to the collection and the historical microenvironment evolution trajectory is calculated to establish a first deviation. The associated microenvironment change trajectory of the associated collection is obtained. A similarity analysis is performed based on the associated microenvironment change trajectory and the microenvironment change trajectory to establish a second deviation. A third consistency score is constructed based on the first deviation and the second deviation. The first consistency score, the second consistency score, and the third consistency score are mapped to the same evaluation space to establish a collection consistency score.

3. The intelligent tracking and inventory method for museum collections as described in claim 2, characterized in that, The dynamic identity fingerprint and micro-environment evolution data of the collections are used to determine their consistency, and collaborative constraints are established for the consistency score of the collections. The consistency determination includes: Based on the microenvironment evolution data, a model for driving environmental changes in the collection is constructed. This model is used to describe the influence of temperature, humidity, light, and vibration on the material response characteristics of the collection. Based on the environmental change-driven model of the collection, the allowable range of material response features in the dynamic identity fingerprint of the collection is adaptively predicted to establish an identity change prediction interval. The currently collected dynamic identity fingerprints of the collections are matched with the identity change prediction interval to establish a collaborative anomaly residual. The collaborative anomaly residuals are introduced as high-priority constraints into the calculation process of the collection consistency score, and consistency calculation constraints are executed.

4. The intelligent tracking and inventory method for museum collections as described in claim 1, characterized in that, Based on the consistency score of the collection, an inventory strategy is triggered to execute intelligent inventory operations, including: The consistency scores of the collections were recorded during multiple inventory cycles to construct a temporal evolution sequence of the consistency scores. Based on the time evolution sequence of the consistency score, the rate of change, fluctuation amplitude and recovery characteristics of the score are extracted to establish an evolution feature set; The evolutionary feature set is compared with the preset stability judgment rules item by item to determine whether the consistency score of the collection meets the conditions of continuous stability, short-term fluctuation or abnormal evolution. When the rate of change of the rating exceeds the preset change threshold, or when the rating fluctuation does not show a regression trend within the preset inventory period, it is judged as an abnormal evolution state. After configuring the intelligent inventory strategy based on the abnormal evolution state, the intelligent inventory operation is executed.

5. The intelligent tracking and inventory method for museum collections as described in claim 1, characterized in that, The three-dimensional geometric contours, surface textures, and material response data of the artifacts are collected to generate multi-dimensional identity feature vectors, including: The key structure decomposition of the collected three-dimensional geometric contour data is performed to extract the set of main contour features used to characterize the overall morphological stability of the collection. Regional statistical processing is performed on surface texture data to extract a set of stable texture features that do not depend on local occlusion and illumination changes; Perform multiple sampling comparisons of material response data to extract a set of material response features that remain consistent under repeated sampling conditions; The contour main feature set, texture stable feature set, and material response feature set are hierarchically combined to form a multidimensional identity feature vector with a fixed feature order and structural representation.

6. The intelligent tracking and inventory method for museum collections as described in claim 5, characterized in that, The collection's dynamic identity fingerprint is established using the multidimensional identity feature vector, including: The multidimensional identity feature vector is used as the initial fingerprint state of the collection's dynamic identity fingerprint; When performing an inventory count, if the preset update conditions are met, an incremental update instruction is generated. The incremental update instruction calls the inventory collection data of the inventory operation and performs incremental amplitude verification of the initial fingerprint state. If the increment magnitude verification passes, incremental update processing is allowed.

7. The intelligent tracking and inventory method for museum collections as described in claim 1, characterized in that, The system utilizes intelligent inventory results and collectible consistency scores to report anomaly risks in collectibles, including: Activate the abnormal risk recording unit, combine the intelligent inventory operation results, the collection consistency score, the abnormal risk judgment results and the risk generation timestamp to establish an abnormal risk evidence data package; The abnormal risk evidence data package is subjected to structured summary processing to generate risk fingerprint data that characterizes the abnormal risk formation process; The risk fingerprint data is written into the block record structure and linked with the preceding abnormal risk record to form an abnormal risk evolution block.

8. The intelligent tracking and inventory method for museum collections as described in claim 1, characterized in that, The system utilizes intelligent inventory results and collectible consistency scores to report anomaly risks in collectibles, including: The results of the intelligent inventory operation and the consistency score of the collection are jointly verified, and the joint verification results are output. Based on the joint verification results, anomaly warning levels are matched, and the anomaly warning level matching results are used to report the abnormal risks of the collection.

9. A smart tracking and inventory system for museum collections, characterized in that: The system for implementing the intelligent tracking and inventory method for museum collections as described in any one of claims 1-8 includes: Fingerprint creation module: Collects the three-dimensional geometric contour, surface texture and material response data of the collection, generates a multi-dimensional identity feature vector, and uses the multi-dimensional identity feature vector to create a dynamic identity fingerprint of the collection; Constraint Construction Module: The exhibition space constraint diagram is constructed by taking the collections, exhibition stands and fixed components in the exhibition space as nodes and establishing edges with relative position constraints, accessibility constraints and synchronous change constraints between nodes. Data acquisition module: Performs microenvironmental data acquisition of the collection, and establishes microenvironmental evolution data including temperature, humidity, light, and vibration data; Joint Analysis Module: Jointly analyzes the dynamic identity fingerprint of the collection, the exhibition space constraint diagram, and the microenvironment evolution data, calculates the residuals with historical data and related collections, and forms a collection consistency score; Intelligent inventory module: Triggers an inventory strategy based on the consistency score of the collection and executes intelligent inventory operations; Risk reporting module: Reports abnormal risks of collections based on the results of intelligent inventory operations and the consistency score of the collections.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the intelligent tracking and inventory method for museum collections as described in any one of claims 1 to 8.

Images