Cultural relic classification method based on cultural relic characteristic curvature

Abstract

The invention relates to the technical field of cultural relic recognition, and provides a cultural relic feature curvature-based cultural relic classification method, which comprises the steps of obtaining feature information of a to-be-recognized cultural relic, and performing cultural relic feature classification; according to cultural relic feature classification, a feature curvature model is established, the feature curvature model is composed of a reference model and an assignment calculation model, the reference model is used for setting a reference curve of each kind of cultural relic features, and the assignment calculation model is used for calculating a comparison deviation curve of the same kind of features of the to-be-recognized cultural relic according to the reference curve; calculating the characteristic curvature of each kind of cultural relic characteristics of the to-be-identified cultural relic according to the comparison deviation curve; and loading the feature curvature into a preset cultural relic classification matching model, and determining a target cultural relic type of the to-be-identified cultural relic. Through the feature curvature model, the classification accuracy can be improved. The reference model and the assignment calculation model can evaluate and compare cultural relic features in a standardized manner, and the influence of subjective judgment is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of cultural relics identification, and in particular to a cultural relics classification method based on characteristic curvature of cultural relics. Background Art

[0002] At present, in the field of cultural relics protection, in order to protect these precious cultural relics, they need to be classified and identified. However, due to the long history of cultural relics and their different preservation conditions, the classification of cultural relics has always been a difficult problem. Traditional classification methods are usually based on manual observation and comparison, but these methods have many limitations, such as being greatly affected by subjective factors and being easily affected by illusions or misjudgments.

[0003] With the development of artificial intelligence, in terms of cultural relics identification, a large database can be built to record the era of cultural relics information, such as historical background information, historical memory information, historical art style information, historical religious information, historical culture information, historical language and characters, totems and other information.

[0004] In 2023, Chungnam National University in South Korea published "Research on Museum Experience Design Based on Sensibility Design-Centered on Relic Image Production in Panoramic Experience Hall", proposing that immersive experience image design provides museum visitors with a new visiting experience, allowing people to immerse themselves in a mysterious virtual environment. This experience allows visitors to transcend reality and enjoy a more interesting exhibition experience. In the form of a digital interactive museum, combined with cultural visual propaganda and interactive cultural images, users can realize the dissemination of cultural relics information based on an emotional experience model. This type of digital museum system is easier to achieve publicity and interactive experience, but there is no specific method for how to classify different cultural relics in digital museums, how to automatically classify them, and how to accurately divide them.

[0005] How to assist archaeologists or digital museums in accurately identifying and reasonably classifying cultural relics through this information? This is a direction that needs to be explored. The existing technology also lacks relevant technologies to assist archaeologists in classifying and identifying cultural relics, which reduces the efficiency of archaeologists and digital museums. Summary of the invention

[0006] The present invention eliminates a cultural relic classification method based on the characteristic curvature of cultural relics, and is used to solve the problem in the prior art that there is a lack of relevant technology to assist archaeologists in classifying and identifying cultural relics, thereby reducing the efficiency of archaeologists.

[0007] The present invention proposes a method for classifying cultural relics based on characteristic curvature of cultural relics, comprising:

[0008] Obtaining feature information of cultural relics to be identified and classifying the features of cultural relics;

[0009] According to the classification of cultural relics, a characteristic curvature model is established, wherein the characteristic curvature model is composed of a reference model and an assignment calculation model. The reference model is used to set a reference curve for each type of cultural relic feature, and the assignment calculation model is used to calculate a comparison deviation curve of the same type of features of the cultural relics to be identified according to the reference curve;

[0010] According to the comparison deviation curve, the characteristic curvature of each type of cultural relic feature of the cultural relic to be identified is calculated;

[0011] The characteristic curvature is loaded into a preset cultural relic classification matching model to determine the target cultural relic type to be identified.

[0012] Furthermore, the characteristic information includes the following types:

[0013] Material characteristics, craft characteristics, cultural characteristics, artistic style characteristics, preservation status characteristics, and historical circulation characteristics.

[0014] Furthermore, the classification of cultural relics characteristics also includes:

[0015] According to the classification type corresponding to the characteristic information, determining that any characteristic information has at least one era dimension of era characteristics; wherein the era characteristics represent the historical and cultural characteristics to which the characteristic information belongs;

[0016] Decomposing the corresponding characteristic information in an information-based manner from at least one era dimension to generate description information of multiple characteristic information;

[0017] Based on the description information, a description database of the characteristic information of the cultural relics to be identified is generated;

[0018] Based on the description database, the characteristic information is classified into cultural relics.

[0019] Furthermore, the classification of cultural relics characteristics also includes:

[0020] Pre-configure cultural relic feature correlation identification rules, wherein the correlation identification rules are configured with identification relationships of strongly correlated cultural relic features;

[0021] The correlation identification rules are used to infer and determine the strongly correlated cultural relic features that must exist in the cultural relic to be identified based on real-time feature information.

[0022] Furthermore, the reference curve includes the following setting steps:

[0023] Obtain the historical time interval of the cultural relic characteristics for which the reference curve is to be set;

[0024] According to the historical time interval, determine the time point when the characteristics of the current cultural relics have changed significantly within the historical time interval;

[0025] According to the significant change time point and the historical time interval, multiple segmentation intervals are set, wherein the endpoint of each segmentation interval corresponds to a significant change time point;

[0026] According to the prominence of the current cultural relic features in each segmented interval, the recognition prominence value is assigned;

[0027] The average value of the significance assignments of the multiple segmentation intervals is determined, and the average value of the significance assignments is compared with the comparison value of each segmentation interval to generate a reference curve.

[0028] Furthermore, the assignment calculation model is also used for:

[0029] Fusing different types of feature information of cultural relics to be identified to determine the initial fusion processing result;

[0030] Determine the spatiotemporal correlation between different feature information of the cultural relics to be identified, and generate spatiotemporal correlation features between different feature information based on the initial fusion processing result and the spatiotemporal correlation between different feature information;

[0031] The initial fusion processing result and the spatiotemporal correlation between different feature information are fused to obtain the target fusion processing result;

[0032] According to the target fusion processing results, determine whether there are irrelevant features and eliminate the irrelevant features.

[0033] Furthermore, the assignment calculation model is used to calculate the comparison deviation curve of the same type of features of the cultural relics to be identified based on the reference curve, and further includes:

[0034] According to the classification of cultural relics characteristics, determine the assignment type and assignment range corresponding to the current characteristic information;

[0035] According to the assignment type, the classification coefficient and the assignment judgment point are set; wherein the assignment judgment point is the assignment attribute of each cultural relic feature;

[0036] Obtain the assigned value of the current cultural relic feature at each assigned determination point, and calculate the assigned mean value;

[0037] When the assignment mean is within the assignment interval, the assignment mean is used as the target assignment of the current cultural relic feature, and the classification coefficient is used as the assignment vector parameter of the current cultural relic feature.

[0038] Furthermore, the step of loading the characteristic curvature into a preset cultural relic classification matching model to determine the target cultural relic type of the cultural relic to be identified includes:

[0039] Construct a differentiated cultural relic matching library, and establish a mapping link between the differentiated cultural relic matching library and the cultural relic classification matching model; wherein the differentiated cultural relic matching library is used for curvature filling, and the mapping link is activated according to the curvature filling, and each mapping link corresponds to a cultural relic textual information;

[0040] Curvature loading is performed through a differentiated artifact matching library, and the corresponding set of artifact types is activated in the artifact matching model through mapping links;

[0041] According to the identification result of the cultural relic type that appears most frequently in the cultural relic type set, the target cultural relic type of the cultural relic to be identified is determined.

[0042] Furthermore, the step of determining the target cultural relic category of the cultural relic to be identified based on the identification result of the cultural relic category that appears most frequently in the cultural relic category set further includes:

[0043] The most frequently appearing cultural relic feature words and the corresponding first cultural relic categories in the cultural relic textual information are calculated. When the cultural relic feature words all belong to the first cultural relic category, the first cultural relic category is used as the target cultural relic category.

[0044] Furthermore, the calculating of the most frequently appearing cultural relic feature words and the corresponding first cultural relic categories in the cultural relic textual information includes:

[0045] Set up statistical calculation units;

[0046] Generate a directed graph for cultural relics classification according to the statistical calculation unit; wherein the nodes in the directed graph for cultural relics classification are used to indicate the characteristic words of cultural relics counted by the statistical calculation unit; and the directed edges in the directed graph for cultural relics classification are used to correspond to different types of cultural relics;

[0047] According to the directed graph of cultural relics classification, the type of cultural relics to which the cultural relics characteristic words belong is determined.

[0048] The beneficial effects of the present invention are:

[0049] The present invention can more accurately identify and classify the characteristics of cultural relics through the characteristic curvature model, thereby improving the accuracy of classification. The reference model and the assignment calculation model provide a standardized method for evaluating and comparing the characteristics of cultural relics, thereby reducing the influence of subjective judgment. The present invention can adapt to cultural relics of different types and styles, and can be applied to a wide range of cultural relic classifications by adjusting the reference curve and the calculation model.

[0050] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0051] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0053] Figure 1 A method flow chart of a method for classifying cultural relics based on characteristic curvature of cultural relics in an embodiment of the present invention;

[0054] Figure 2 A schematic diagram of a classification diagram of cultural relics characteristics in an embodiment of the present invention;

[0055] Figure 3 It is a schematic diagram of implementing the classification of cultural relics characteristics based on characteristic information in an embodiment of the present invention;

[0056] Figure 4 A schematic diagram of an implementation of an embodiment of the present invention applied in a digital museum;

[0057] Figure 5 A schematic diagram of a three-dimensional display of an embodiment of the present invention applied in a digital museum;

[0058] Figure 6 This is a curvature comparison diagram of different types of cultural relics applied in the feature assignment of cultural relics classification according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0060] like Figure 1 As shown, this embodiment provides a method for classifying cultural relics based on characteristic curvature of cultural relics, including:

[0061] Obtaining feature information of cultural relics to be identified and classifying the features of cultural relics;

[0062] According to the classification of cultural relics, a characteristic curvature model is established, wherein the characteristic curvature model is composed of a reference model and an assignment calculation model. The reference model is used to set a reference curve for each type of cultural relic feature, and the assignment calculation model is used to calculate a comparison deviation curve of the same type of features of the cultural relics to be identified according to the reference curve;

[0063] According to the comparison deviation curve, the characteristic curvature of each type of cultural relic feature of the cultural relic to be identified is calculated;

[0064] The characteristic curvature is loaded into a preset cultural relic classification matching model to determine the target cultural relic type to be identified.

[0065] The principle of the above technical solution is:

[0066] like Figure 1 As shown, the function of the present invention is mainly to assist in the classification and identification of cultural relics, and is applied to the archaeological field and the automatic update of the cultural relics database of digital museums. In the specific implementation process, the characteristic information of the cultural relics to be identified is first obtained for analysis, and the characteristics of the cultural relics are classified according to the characteristic information of the cultural relics. In the specific implementation process, the characteristic information of the cultural relics to be identified is extracted through intelligent vision and the information uploaded by the cultural relics uploader. In the prior art, for digital museums, when the administrator uploads the cultural relics information, the information of the cultural relics may be inaccurate or incomplete. The present invention can adopt visual feature acquisition technology; for example: high-resolution cameras, measuring instruments, multi-spectral technology, etc. Then, the color information of the cultural relics can be recorded by high-fidelity color shooting; different lighting and shooting angles are used to highlight the texture details of the cultural relics. The three-dimensional shape of the cultural relics is obtained by three-dimensional scanning or multi-view stereo matching technology; the tiny details and surface conditions of the cultural relics are captured using macro photography technology.

[0067] When the present invention is implemented, various characteristic information of cultural relics, such as color, shape, texture, etc., are first collected through the above-mentioned visual technology, and then these characteristic information are clustered and classified. The clustering and classification process uses a general convolutional neural network. The convolutional neural network is trained with cultural relics data to generate a cultural relics feature recognition network. After the characteristic information of the cultural relics to be identified is input into the cultural relics feature recognition network, the characteristic classification result of the cultural relics is determined. According to the characteristic classification results, different characteristics of cultural relics can be identified, and cultural relics with incorrect characteristic data uploaded can be judged. For example, if the historical period and historical story of the cultural relics are wrong, it can be judged that the cultural relics information is wrong when it is judged that the characteristics of the cultural relics do not belong to this period or do not belong to this story.

[0068] The core of the present invention is to establish a characteristic curvature model according to the characteristic classification results of cultural relics. The characteristic curvature model is composed of a reference model and an assignment calculation model, and both the reference model and the assignment calculation model can realize the calculation and analysis of the characteristic values ​​of cultural relics. Among them, the reference model is used to set the reference curve of each type of cultural relic characteristics (for example: the painting style pattern characteristics of blue and white porcelain cultural relics are characterized by cobalt blue patterns, the color value when fired at 1250 degrees Celsius is used as the first reference parameter, and the body body whiteness degree in the LAB color space is 100 as the reference parameter; in addition, the reference parameters of glaze characteristics, the prosperous time points of different crafts and techniques in the production process can also be set as reference parameters, etc.; these parameters form a curve, the ordinate of the curve is the feature; the abscissa of the curve is the year and month), the assignment calculation model is used to calculate the comparison deviation curve of the same type of characteristics of the cultural relics to be identified according to the reference curve, and the assignment calculation model is also used to assign values ​​to the cultural relic characteristics. The construction of the reference model and the assignment calculation model is based on the assignment recognition technology, and the characteristic information of the cultural relics is extracted and analyzed according to the parameters obtained by the assignment.

[0069] In the recognition process, the present invention calculates the characteristic curvature of each type of cultural relic feature of the cultural relic to be recognized according to the comparison deviation curve. The unique curvature of different cultural relics can be determined by the calculation formula of the curvature. The calculation formula of the curvature is based on the characteristic curvature model, and the characteristic curvature of the cultural relic to be recognized is obtained by performing an integral operation on the characteristic data of the cultural relic to be recognized.

[0070] Finally, the present invention loads the characteristic curvature into the preset cultural relic classification matching model, and determines the target cultural relic type of the object to be identified. In the above process, the characteristic data of the cultural relic to be identified and the cultural relic classification matching model are matched through the matching algorithm, so as to obtain the target cultural relic type of the object to be identified, that is, the cultural relic identification is realized based on the curvature. In the process of cultural relic identification, the value is unique and the identification result is more accurate. Moreover, the horizontal coordinate in the characteristic curvature model is the year and month, so it is infinite. Therefore, when calculating the characteristics of each cultural relic, the curvature must be inconsistent. As shown in the attached figure Figure 6 As shown, in the process of feature assignment, the appearance time of the cultural relic features and the assigned values ​​of different features can determine the curvature of different cultural relics, and then each cultural relic can be classified according to the reference curve.

[0071] In the specific implementation process, the present invention can also be combined with an immersive experience. In a digital museum, based on automatic classification of cultural relics, the division and classification of different cultural relics products in the digital museum can be realized, so as to quickly retrieve the corresponding cultural relic information. In the process of cultural relic identification, cultural relic analysis is carried out. The present invention can be combined with the image design of the museum. When the user is in an immersive cultural experience process, the characteristics of different cultural relics can be analyzed and identified with the help of feature recognition technology and curvature calculation technology. Through specific curvature parameters, the corresponding database is called, and combined with image recognition, the image of the cultural relic to be identified is displayed in three dimensions, and an interactive experience is carried out with the user through the different characteristics of the text and image to be identified. As shown in the attached figure Figure 4 As shown in the figure, when interacting with cultural relics in a digital museum, the cultural relics can be introduced and visually promoted through three-dimensional display forms and brief cultural relic descriptions. Figure 5 , the specific information and layout of the cultural relics can be three-dimensionally mapped in detail, so as to achieve accurate analysis of different characteristics, and accurate calling and classification of different cultural relic characteristics.

[0072] The beneficial effects of the above technical solution are:

[0073] The present invention can more accurately identify and classify the characteristics of cultural relics through the characteristic curvature model, thereby improving the accuracy of classification. The reference model and the assignment calculation model provide a standardized method for evaluating and comparing the characteristics of cultural relics, thereby reducing the influence of subjective judgment. The present invention can adapt to cultural relics of different types and styles, and can be applied to a wide range of cultural relic classifications and digital cultural relic updates by adjusting the reference curve and the calculation model.

[0074] As an embodiment of the present invention, the feature information includes the following types:

[0075] Material characteristics, craft characteristics, cultural characteristics, artistic style characteristics, preservation status characteristics, and historical circulation characteristics.

[0076] The principle of the above technical solution is:

[0077] like Figure 2 As shown in the figure, this is a classification diagram of cultural relics characteristics, showing the six major categories and their subcategories of cultural relics characteristics. This classification diagram helps us to more comprehensively understand and analyze the various attribute characteristics of cultural relics, so as to achieve more accurate classification and categorization. Specific subcategories are listed under each major category (such as material characteristics, process characteristics, etc.), for example, material characteristics include material, texture, density and hardness, etc. This classification method helps to take into account multiple aspects of information in the identification and research of cultural relics. When it is implemented:

[0078] Material characteristics include but are not limited to: material, texture, density, hardness, etc.;

[0079] Craft features include but are not limited to: production process, processing method, decoration technique, etc.;

[0080] Cultural characteristics include but are not limited to: religious beliefs, regional characteristics, historical background, etc.;

[0081] Artistic style features include but are not limited to: painting techniques, composition features, color matching, etc.;

[0082] Preservation characteristics include, but are not limited to: degree of damage, cracks, corrosion, etc.;

[0083] Historical circulation characteristics include but are not limited to: collection time, circulation experience, original owner and other information.

[0084] The beneficial effects of the above technical solution are:

[0085] It is convenient to comprehensively analyze the various attribute characteristics of cultural relics and realize accurate classification and categorization. In addition, by combining the method of cultural relics feature classification, more refined feature collection and feature assignment of cultural relics can be realized.

[0086] As an embodiment of the present invention, the classification of cultural relics characteristics further includes:

[0087] According to the classification type corresponding to the characteristic information, determining that any characteristic information has at least one era dimension of era characteristics; wherein the era characteristics represent the historical and cultural characteristics to which the characteristic information belongs;

[0088] Decomposing the corresponding characteristic information in an information-based manner from at least one era dimension to generate description information of multiple characteristic information;

[0089] Based on the description information, a description database of the characteristic information of the cultural relics to be identified is generated;

[0090] Based on the description database, the characteristic information is classified into cultural relics.

[0091] The principle of the above technical solution is:

[0092] like Figure 3As shown, the present invention determines at least one era dimension of era characteristics of any feature information according to the classification type corresponding to the feature information; wherein the era characteristics represent the historical and cultural characteristics to which the feature information belongs. This is to summarize and conclude different types of feature information through literature research, expert interviews, etc., so as to determine the period and cultural background represented by each feature information. The corresponding feature information is decomposed from at least one era dimension to generate description information of multiple feature information. This is to describe the feature information in a textual manner through natural language processing, machine learning and other technical means to obtain multiple description feature information. These description feature information may include the specific description of the feature information, the correlation with other feature information, the degree of influence on the value of cultural relics, and other information. Based on the description information, a description database of the feature information of cultural relics to be identified is generated. Through data integration, cleaning, deduplication and other operations, the description information of the generated multiple feature information is integrated into a database, and the description information is coded, indexed, and other processes are performed on it, so as to facilitate the programmatic processing of cultural relics characteristics. According to the description database, the feature information is classified into cultural relics characteristics. This is to classify and categorize the feature information through a rule-based method or a machine learning algorithm to obtain the final classification result of cultural relics characteristics. This can provide reference and basis for the fields of cultural relics protection, research, and display.

[0093] The beneficial effects of the above technical solution are:

[0094] The present invention can analyze the characteristics of cultural relics, determine the description information of the characteristics of cultural relics, and accurately classify the characteristics of cultural relics through information decomposition of at least one era dimension.

[0095] As an embodiment of the present invention, the classification of cultural relics characteristics further includes:

[0096] Pre-configure cultural relic feature correlation identification rules, wherein the correlation identification rules are configured with identification relationships of strongly correlated cultural relic features;

[0097] The correlation identification rules are used to infer and determine the strongly correlated cultural relic features that must exist in the cultural relic to be identified based on real-time feature information.

[0098] The principle of the above technical solution is:

[0099] The present invention pre-configures the cultural relic feature correlation identification rules, wherein the correlation identification rules are configured with the identification relationship of the strongly correlated cultural relic features. The strongly correlated cultural relic feature identification relationship is to pre-define the correlation relationship between some common cultural relic features through literature research, expert interviews, etc., combined with the actual needs of cultural relic protection, and convert it into a correlation identification rule, wherein the identification relationship of the strongly correlated cultural relic features refers to the existence of a strong positive correlation between two feature information, which can be used to determine whether the two feature information are the same type of cultural relic, or that if there is A, there must be B between the two features, so as to determine whether there is erroneous feature information. The present invention uses the correlation identification rules based on real-time feature information to infer the strongly correlated cultural relic features that must exist in the cultural relic to be identified. Combined with the feature identification technology in the prior art, the feature information collected in real time is analyzed, and compared with the feature information in the existing feature database, so as to find the correlation relationship between the two feature information. If there is a strong correlation between the two feature information, it can be considered that the two feature information are feature information related to the same type of cultural relic, so that the cultural relic to be identified can be classified into the corresponding category by feature matching.

[0100] The beneficial effects of the above technical solution are:

[0101] The present invention can determine the correlation between the features of different cultural relics, thereby determining whether there are misidentified features or features of the same type of cultural relics, thereby enabling rapid identification of cultural relics.

[0102] As an embodiment of the present invention: the reference curve includes the following setting steps:

[0103] Obtain the historical time interval of the cultural relic characteristics for which the reference curve is to be set;

[0104] According to the historical time interval, determine the time point when the characteristics of the current cultural relics have changed significantly within the historical time interval;

[0105] According to the significant change time point and the historical time interval, multiple segmentation intervals are set, wherein the endpoint of each segmentation interval corresponds to a significant change time point;

[0106] According to the prominence of the current cultural relic features in each segmented interval, the recognition prominence value is assigned;

[0107] The average value of the significance assignments of the multiple segmentation intervals is determined, and the average value of the significance assignments is compared with the comparison value of each segmentation interval to generate a reference curve.

[0108] The principle of the above technical solution is:

[0109] During the implementation of the present invention, the historical time interval of the cultural relic features of the reference curve to be set will be obtained, which is obtained by collecting a certain number of cultural relic feature data and sorting them in chronological order, thereby obtaining the historical time interval of the cultural relic features. According to the historical time interval, the significant change time point of the current cultural relic features within the historical time interval is determined; this is to analyze the time series of cultural relic features through technical means such as time series analysis and cluster analysis, find the significant change point therein, and thus determine the significant change time point. According to the significant change time point and the historical time interval, multiple segmentation intervals are set, wherein the endpoint of each segmentation interval corresponds to a significant change time point; this is to obtain a series of segmentation intervals by dividing the historical time interval, wherein the endpoint of each segmentation interval corresponds to a significant change time point. According to the significance of the current cultural relic features in each segmentation interval, identification significance assignment is performed; by calculating the significance of the current cultural relic features in each segmentation interval, different identification significance assignments are assigned to it. Determine the average value of the significance assignments of multiple segmentation intervals, and generate a reference curve by comparing the average value of the significance assignments with the comparison value of each segmentation interval; calculate the average value of the significance assignments in each segmentation interval by mathematical methods such as averaging and least squares method, and use the average value as the slope parameter of the reference curve, and generate a reference curve based on the comparison value in each segmentation interval.

[0110] The beneficial effects of the above technical solution are:

[0111] The present invention can set reference curves of different cultural relic features according to the significant features of the cultural relics, so as to achieve significant comparative reference when determining the cultural relic features.

[0112] As an embodiment of the present invention: the assignment calculation model is also used for:

[0113] Fusing different types of feature information of cultural relics to be identified to determine the initial fusion processing result;

[0114] Determine the spatiotemporal correlation between different feature information of the cultural relics to be identified, and generate spatiotemporal correlation features between different feature information based on the initial fusion processing result and the spatiotemporal correlation between different feature information;

[0115] The initial fusion processing result and the spatiotemporal correlation between different feature information are fused to obtain the target fusion processing result;

[0116] According to the target fusion processing results, determine whether there are irrelevant features and eliminate the irrelevant features.

[0117] The principle of the above technical solution is:

[0118] The present invention fuses the different types of feature information of the cultural relics to be identified to determine the initial fusion processing result; this is to fuse the feature information of different types of cultural relics, for example: the glaze characteristics and glaze craftsmanship of blue and white porcelain are fused to obtain the initial fusion processing result, after the fusion, there will be the correlation result of the glaze and glaze craftsmanship of blue and white porcelain, the overlapping historical time of the craftsmanship and glaze characteristics of the historical period of the cultural relics is carried out, and the historical time point of the cultural relics is located. Of course, after the fusion of the two, there may be no overlapping time, which means that it is very likely that the blue and white porcelain is a fake made by later generations or there is a historical overlap fault period.

[0119] Determine the spatiotemporal correlation between different feature information of the cultural relics to be identified, and generate spatiotemporal correlation features between different feature information based on the initial fusion processing results and the spatiotemporal correlation between different feature information; in specific implementation, determine whether there is overlap in time and space by judging the construction correlation, that is, the correlation between the glaze features and the glaze craftsmanship in time and space in the above embodiment. Then, if these two features have spatiotemporal correlation or not, combine them with other cultural relic features respectively, and judge the spatiotemporal correlation features respectively to obtain many fusion results. Then, based on many fusion results, eliminate the fusion results of cultural relic features that are completely impossible to overlap, that is, appearing in the same historical period and historical location.

[0120] The beneficial effects of the above technical solution are:

[0121] The present invention can determine whether there are unrelated features during the feature recognition process, thereby deleting the corresponding features, making the recognition results of the cultural relics features more accurate.

[0122] As an embodiment of the present invention: the assignment calculation model is used to calculate the comparison deviation curve of the same type of features of the cultural relics to be identified according to the reference curve, and further includes:

[0123] According to the classification of cultural relics characteristics, determine the assignment type and assignment range corresponding to the current characteristic information;

[0124] According to the assignment type, the classification coefficient and the assignment judgment point are set; wherein the assignment judgment point is the assignment attribute of each cultural relic feature;

[0125] Obtain the assigned value of the current cultural relic feature at each assigned determination point, and calculate the assigned mean value;

[0126] When the assignment mean is within the assignment interval, the assignment mean is used as the target assignment of the current cultural relic feature, and the classification coefficient is used as the assignment vector parameter of the current cultural relic feature.

[0127] The principle of the above technical solution is:

[0128] The present invention classifies cultural relics according to their characteristics and determines the assignment type and assignment interval corresponding to the current characteristic information; for example, the painting technique of the pattern on the blue and white porcelain is taken as a time-space interval, and the assignment interval determines the maximum assignment, the minimum assignment and the intermediate assignment of different time-space nodes of the pattern on the blue and white porcelain within this time-space interval.

[0129] According to the assignment type, the assignment classification coefficient and assignment judgment point are set; wherein the assignment judgment point is the assignment attribute of each cultural relic feature; the assignment type can determine the feature type, the classification coefficient is the type coefficient pre-set in the corresponding feature type of the cultural relic feature, and the assignment judgment point is based on the assignment attribute, namely material feature, process feature, etc.

[0130] The value of the current cultural relic feature is assigned at each assignment decision point, and the assignment mean is calculated; this is achieved through database query, data capture and other technical means to obtain the value of the current cultural relic feature corresponding to each assignment decision point, and calculate the corresponding assignment mean.

[0131] When the assignment mean is within the assignment interval, the assignment mean is used as the target assignment of the current cultural relic feature, and the classification coefficient is used as the assignment vector parameter of the current cultural relic feature; this is to calculate the target assignment of the current cultural relic feature based on the assignment mean of the current feature information and the corresponding classification coefficient, and use the classification coefficient as the assignment vector parameter of the current cultural relic feature.

[0132] The beneficial effects of the above technical solution are:

[0133] The present invention can set an assignment interval, so that in a specific assignment calculation process, a small difference in the assignment can be used to make a larger difference in the recognition result, thereby preventing recognition errors.

[0134] As an embodiment of the present invention, the step of loading the characteristic curvature into a preset cultural relic classification matching model to determine the target cultural relic type to be identified includes:

[0135] Construct a differentiated cultural relic matching library, and establish a mapping link between the differentiated cultural relic matching library and the cultural relic classification matching model; wherein the differentiated cultural relic matching library is used for curvature filling, and the mapping link is activated according to the curvature filling, and each mapping link corresponds to a cultural relic textual information;

[0136] Curvature loading is performed through a differentiated artifact matching library, and the corresponding set of artifact types is activated in the artifact matching model through mapping links;

[0137] According to the identification result of the cultural relic type that appears most frequently in the cultural relic type set, the target cultural relic type of the cultural relic to be identified is determined.

[0138] The principle of the above technical solution is:

[0139] The present invention constructs a differentiated cultural relic matching library, and establishes a mapping link between the differentiated cultural relic matching library and the cultural relic classification matching model; wherein the differentiated cultural relic matching library is used for curvature filling, and the mapping link is activated according to the curvature filling, and each mapping link corresponds to a cultural relic textual information; this is achieved by constructing a differentiated cultural relic matching library (a matching library of curvature difference and cultural relic type), and mapping the cultural relic matching model and the cultural relic classification matching model according to the curvature filling situation (curvature intersection situation) in the matching library to form a mapping link.

[0140] Curvature is loaded through the differentiated cultural relic matching library, and the corresponding cultural relic type set is activated in the cultural relic matching model through mapping links; this is to extract the feature data of the cultural relic to be identified from the cultural relic matching model, and activate it through mapping links, and output the activated cultural relic type set as the final cultural relic classification result.

[0141] According to the identification result of the cultural relic type that appears most frequently in the cultural relic type set, the target cultural relic type of the cultural relic to be identified is determined; this is achieved by counting the frequency of occurrence of each cultural relic type in the cultural relic type set, and selecting the cultural relic type with the highest frequency of occurrence as the final target cultural relic type.

[0142] The beneficial effects of the above technical solution are:

[0143] The present invention determines the type of target cultural relic by the number of times the identification of the cultural relic type occurs, and based on a strongly correlated identification method, determines the type of cultural relic more accurately, and the identification result is more accurate.

[0144] As an embodiment of the present invention, the step of determining the target cultural relic type to be identified according to the identification result of the cultural relic type that appears most frequently in the cultural relic type set further includes:

[0145] The most frequently appearing cultural relic feature words and the corresponding first cultural relic categories in the cultural relic textual information are calculated. When the cultural relic feature words all belong to the first cultural relic category, the first cultural relic category is used as the target cultural relic category.

[0146] The principle of the above technical solution is:

[0147] The present invention calculates the most frequently appearing cultural relic feature words and the corresponding first cultural relic category in the cultural relic textual information. When the cultural relic feature words all belong to the first cultural relic category, the first cultural relic category is used as the target cultural relic category. The most frequently appearing words and their corresponding categories in the textual information of the cultural relic to be identified are calculated, and the first cultural relic category with the highest frequency of occurrence is used as the target cultural relic category. If all the feature words of the cultural relic to be identified belong to the same cultural relic category, the category is used as the final target cultural relic category.

[0148] As an embodiment of the present invention, the step of calculating the most frequently appearing cultural relic feature words and the corresponding first cultural relic categories in the cultural relic textual information includes:

[0149] Set up statistical calculation units;

[0150] Generate a directed graph for cultural relics classification according to the statistical calculation unit; wherein the nodes in the directed graph for cultural relics classification are used to indicate the characteristic words of cultural relics counted by the statistical calculation unit; and the directed edges in the directed graph for cultural relics classification are used to correspond to different types of cultural relics;

[0151] According to the directed graph of cultural relics classification, the type of cultural relics to which the cultural relics characteristic words belong is determined.

[0152] The principle of the above technical solution is:

[0153] The present invention selects one or more statistical calculation units in the cultural relic feature information, and counts the cultural relic feature words in each statistical calculation unit, determines the number of times different feature words appear in the statistical calculation unit, determines the feature word with the largest number of appearances, and records it as the first cultural relic category. According to the combination of the first cultural relic category and the cultural relic feature words, a cultural relic classification directed graph is generated, in which the nodes represent the cultural relic feature words in each statistical calculation unit, and the directed edges represent the classification relationship of different types of cultural relics.

[0154] Traverse all nodes and directed edges in the cultural relic classification directed graph, and determine the cultural relic category to which the feature word represented by each node belongs based on the statistical information of the node and the type information of the directed edge. If all the feature words in a statistical calculation unit belong to the same cultural relic category, then this category is used as the target cultural relic category.

[0155] The beneficial effects of the above technical solution are:

[0156] The present invention can automatically classify and identify cultural relics from the perspective of textualization and visualization, thereby improving the accuracy and efficiency of cultural relic classification.

[0157] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for classifying cultural relics based on characteristic curvature of cultural relics, characterized in that: include: Obtaining feature information of cultural relics to be identified and classifying the features of cultural relics; According to the classification of cultural relics, a characteristic curvature model is established, wherein the characteristic curvature model is composed of a reference model and an assignment calculation model. The reference model is used to set a reference curve for each type of cultural relic feature, and the assignment calculation model is used to calculate a comparison deviation curve of the same type of features of the cultural relics to be identified according to the reference curve; According to the comparison deviation curve, the characteristic curvature of each type of cultural relic feature of the cultural relic to be identified is calculated; The characteristic curvature is loaded into a preset cultural relic classification matching model to determine the target cultural relic type to be identified.

2. A method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The characteristic information includes the following types: Material characteristics, craft characteristics, cultural characteristics, artistic style characteristics, preservation status characteristics, and historical circulation characteristics.

3. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The classification of cultural relics characteristics also includes: According to the classification type corresponding to the characteristic information, determining that any characteristic information has at least one era dimension of era characteristics; wherein the era characteristics represent the historical and cultural characteristics to which the characteristic information belongs; Decomposing the corresponding characteristic information in an information-based manner from at least one era dimension to generate description information of multiple characteristic information; Based on the description information, a description database of the characteristic information of the cultural relics to be identified is generated; Based on the description database, the characteristic information is classified into cultural relics.

4. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The classification of cultural relics characteristics also includes: Pre-configure cultural relic feature correlation identification rules, wherein the correlation identification rules are configured with identification relationships of strongly correlated cultural relic features; The correlation identification rules are used to infer and determine the strongly correlated cultural relic features that must exist in the cultural relic to be identified based on real-time feature information.

5. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The reference curve includes the following setting steps: Obtain the historical time interval of the cultural relic characteristics for which the reference curve is to be set; According to the historical time interval, determine the time point when the characteristics of the current cultural relics have changed significantly within the historical time interval; According to the significant change time point and the historical time interval, multiple segmentation intervals are set, wherein the endpoint of each segmentation interval corresponds to a significant change time point; According to the prominence of the current cultural relic features in each segmented interval, the recognition prominence value is assigned; The average value of the significance assignments of the multiple segmentation intervals is determined, and the average value of the significance assignments is compared with the comparison value of each segmentation interval to generate a reference curve.

6. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The assignment calculation model is also used for: Fusing different types of feature information of cultural relics to be identified to determine the initial fusion processing result; Determine the spatiotemporal correlation between different feature information of the cultural relics to be identified, and generate spatiotemporal correlation features between different feature information based on the initial fusion processing result and the spatiotemporal correlation between different feature information; The initial fusion processing result and the spatiotemporal correlation between different feature information are fused to obtain the target fusion processing result; According to the target fusion processing results, determine whether there are irrelevant features and eliminate the irrelevant features.

7. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The assignment calculation model is used to calculate the comparison deviation curve of the same type of features of the cultural relics to be identified according to the reference curve, and further includes: According to the classification of cultural relics characteristics, determine the assignment type and assignment range corresponding to the current characteristic information; According to the assignment type, the classification coefficient and the assignment judgment point are set; wherein the assignment judgment point is the assignment attribute of each cultural relic feature; Obtain the assigned value of the current cultural relic feature at each assigned determination point, and calculate the assigned mean value; When the assignment mean is within the assignment interval, the assignment mean is used as the target assignment of the current cultural relic feature, and the classification coefficient is used as the assignment vector parameter of the current cultural relic feature.

8. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 1, characterized in that: The step of loading the characteristic curvature into a preset cultural relic classification matching model to determine the target cultural relic type to be identified includes: Construct a differentiated cultural relic matching library, and establish a mapping link between the differentiated cultural relic matching library and the cultural relic classification matching model; wherein the differentiated cultural relic matching library is used for curvature filling, and the mapping link is activated according to the curvature filling, and each mapping link corresponds to a cultural relic textual information; Curvature loading is performed through a differentiated artifact matching library, and the corresponding set of artifact types is activated in the artifact matching model through mapping links; According to the identification result of the cultural relic type that appears most frequently in the cultural relic type set, the target cultural relic type of the cultural relic to be identified is determined.

9. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 8, characterized in that: The step of determining the target cultural relic category of the cultural relic to be identified based on the identification result of the cultural relic category that appears most frequently in the cultural relic category set also includes: The most frequently appearing cultural relic feature words and the corresponding first cultural relic categories in the cultural relic textual information are calculated. When the cultural relic feature words all belong to the first cultural relic category, the first cultural relic category is used as the target cultural relic category.

10. The method for classifying cultural relics based on characteristic curvature of cultural relics as claimed in claim 8, characterized in that: The calculating of the most frequently appearing cultural relic feature words and the corresponding first cultural relic categories in the cultural relic textual information includes: Set up statistical calculation units; Generate a directed graph for cultural relics classification according to the statistical calculation unit; wherein the nodes in the directed graph for cultural relics classification are used to indicate the characteristic words of cultural relics counted by the statistical calculation unit; and the directed edges in the directed graph for cultural relics classification are used to correspond to different types of cultural relics; According to the directed graph of cultural relics classification, the type of cultural relics to which the cultural relics characteristic words belong is determined.

Images