Intelligent cultural symbol identification and analysis method and system based on image semantic understanding

Abstract

The invention provides a cultural symbol intelligent identification and analysis method and system based on image semantic understanding, and belongs to the technical field of semantic identification. The method comprises the steps of 1, collecting an original image, and obtaining image features of the original image based on an image feature analysis algorithm; 2, performing semantic analysis on the image features, and identifying and extracting national culture symbol information; 3, generating direction index score values of spatial mutual embedding degree, economic co-richness, culture fusion degree, social harmony degree and psychological recognition degree based on the national culture symbol information; 4, fusing all direction index score values to generate a national harmony index of the original image; and 5, based on a visualization technology, combining the national harmony index and the direction index score value of the original image to generate a national communication evaluation result. Through image feature analysis and semantic recognition, the national harmony index is evaluated, and the cultural symbol recognition efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of semantic recognition, and particularly to a method and system for intelligent recognition and analysis of cultural symbols based on image semantic understanding. Background Art

[0002] Currently, with the acceleration of the globalization process, cultural diversity has become an important feature of social development. As an important carrier of national culture, cultural symbols carry rich historical backgrounds, traditional values, and social identities. However, traditional image recognition often can only simply recognize the ethnic visual features of images, but cannot understand the complex semantic backgrounds in these images.

[0003] Therefore, the present invention proposes a method and system for intelligent recognition and analysis of cultural symbols based on image semantic understanding. Summary of the Invention

[0004] The present invention provides a method and system for intelligent recognition and analysis of cultural symbols based on image semantic understanding, which is used to identify ethnic cultural symbols and extract relevant information through image feature analysis and semantic understanding, generate scores of multiple social and cultural indicators, and finally fuse them into an ethnic harmony index, and use visualization technology to present the evaluation results of ethnic exchanges, so as to improve the efficiency of cultural symbol recognition.

[0005] On the one hand, the present invention provides a method for intelligent recognition and analysis of cultural symbols based on image semantic understanding, including:

[0006] Step 1: Collect the original image, and obtain the image features of the original image based on the image feature analysis algorithm;

[0007] Step 2: Perform semantic analysis on the image features to identify and extract ethnic cultural symbol information;

[0008] Step 3: Generate directional index score values of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological identity degree according to the ethnic cultural symbol information;

[0009] Step 4: Fuse all the directional index score values to generate the ethnic harmony index of the original image;

[0010] Step 5: Based on the visualization technology, combine the ethnic harmony index and the directional index score values of the original image to generate the evaluation results of ethnic exchanges.

[0011] On the other hand, after collecting the original image, it includes:

[0012] Collect the original image as the analysis target according to the preset target data source;

[0013] Perform gray-scale processing on each pixel of the original image to generate a first image:

[0014] Use a Gaussian filter kernel to perform smoothing processing on the first image, take the weighted average of the pixels in its neighborhood for each pixel, and then perform smoothing processing on the pixel according to the weighted average in combination with the Gaussian filter kernel to obtain a second image after filtering to remove noise.

[0015] On the other hand, obtain the image features of the original image based on the image feature analysis algorithm, including:

[0016] Construct a horizontal direction operator and a vertical direction operator according to a preset template, and calculate the gradient magnitude of any pixel in the second image based on the horizontal direction operator and the vertical direction operator;

[0017] Take any pixel point in the second image as the central pixel point, obtain the surrounding pixel points of the central pixel point according to a preset distance, and calculate the neighborhood pixel gradient magnitude difference matrix between the central pixel point and any surrounding pixel point, and statistically calculate the neighborhood pixel gradient magnitude difference matrices of all pixel points;

[0018] Input the neighborhood pixel gradient magnitude difference matrix of any pixel point into a preset response function and output a response value. If the response value is greater than a preset threshold, determine that the pixel point is an edge point;

[0019] All edge points of the second image constitute the image features of the second image.

[0020] On the other hand, perform semantic analysis on the image features to identify and extract ethnic cultural symbol information, including:

[0021] Obtain the expert analysis results of ethnic cultural symbols and construct templates of multiple cultural symbol semantic types;

[0022] Slide and match any template on the image features, and generate a sliding match similarity each time it slides;

[0023] Obtain the maximum sliding match similarity between the template and the image features according to a preset maximum value function. If the maximum sliding match similarity is greater than a preset standard match threshold, determine that the image features have the cultural symbol semantic type corresponding to the template, otherwise it means it does not exist;

[0024] According to the matching results of the image features and all templates, obtain all matching cultural symbol semantic types and the maximum sliding match similarity, which constitute the ethnic cultural symbol information of the image features.

[0025] On the other hand, determine the subordinate indicators of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree, including:

[0026] Generate corresponding secondary indicators according to all semantic types of cultural symbols;

[0027] Using a word embedding model, input all secondary indicators and five preset directional indicators, including: spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological identity degree, to generate corresponding semantic vector values;

[0028] Taking any directional indicator as the center for clustering, obtaining the semantic similarity between any secondary indicator and all center clusters, and determining the center cluster with the highest semantic similarity as the directional indicator to which the secondary indicator belongs;

[0029] According to all secondary indicators corresponding to any directional indicator, determining the semantic similarity between any two secondary indicators, classifying those with a semantic similarity greater than the first threshold into the same unit class, generating multiple unit classes under the directional indicator, and naming any unit class according to the semantic characteristics of the unit class to obtain the first-level indicator;

[0030] Conduct a questionnaire survey on a preset group through social identity theory to obtain the voting evaluations of the first-level indicators, secondary indicators, and directional indicators, and quantify the voting evaluations by combining data statistical methods to obtain the weights of the first-level indicators, secondary indicators, and directional indicators, and obtain the standard values of the secondary indicators from national data sources.

[0031] On the other hand, generate the directional indicator score values of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological identity degree, including:

[0032] Based on the ethnic cultural symbol information of image features, calculate the score value of any first-level indicator of the image features;

[0033] Based on the score values of all first-level indicators and their weights, generate the directional indicator score values of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological identity degree.

[0034] On the other hand, fuse all the directional indicator score values to generate the ethnic harmony index of the original image, including:

[0035] Based on the five directional indicators and their weights, perform weighted fusion calculation according to the directional indicator score values to generate the ethnic harmony index of the original image.

[0036] On the other hand, based on visualization technology, generate the ethnic communication evaluation result of the original image, including:

[0037] Based on visualization technology, the ethnic harmony index of the original image and the score values of all direction indicators are used as the main display part of the visualization report; any direction indicator and its corresponding score value are used as the subordinate display part. Select any direction indicator dropdown to jointly display all first-level indicators - second-level indicators, and construct a multi-level visualization page for intelligent recognition and analysis of cultural symbols.

[0038] Based on the blank report template of the multi-level visualization page for intelligent recognition and analysis of cultural symbols, input the second-level indicators of the original image and their score values, and output the ethnic communication evaluation result of the original image.

[0039] On the other hand, a cultural symbol intelligent recognition and analysis system based on image semantic understanding is applied to the above-mentioned cultural symbol intelligent recognition and analysis method based on image semantic understanding, including:

[0040] Image feature module: Collect the original image and obtain the image features of the original image based on the image feature analysis algorithm.

[0041] Extraction module: Perform semantic analysis on the image features to identify and extract ethnic cultural symbol information.

[0042] Direction indicator module: Generate score values of direction indicators such as spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree according to the ethnic cultural symbol information.

[0043] Ethnic harmony index module: Integrate all direction indicator score values to generate the ethnic harmony index of the original image.

[0044] Visualization module: Based on visualization technology, combine the ethnic harmony index and direction indicator score values of the original image to generate the ethnic communication evaluation result.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] The present invention provides a cultural symbol intelligent recognition and analysis method and system based on image semantic understanding, which improves the accuracy of cultural symbol image recognition, the ability of cultural context understanding, and efficiency through deep learning and cross-modal fusion technology, is applicable to multi-cultural environments, supports automated and large-scale processing, and promotes the development of cultural heritage protection and intelligent applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0048] Figure 1 It is a schematic flowchart of a method for intelligent recognition and analysis of cultural symbols based on image semantic understanding provided by an embodiment of the present invention;

[0049] Figure 2 It is a schematic structural diagram of a system for intelligent recognition and analysis of cultural symbols based on image semantic understanding provided by an embodiment of the present invention. Detailed implementation manners

[0050] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without making creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0051] Embodiment 1:

[0052] As Figure 1 shown, the method for intelligent recognition and analysis of cultural symbols based on image semantic understanding provided by an embodiment of the present invention includes:

[0053] Step 1: Collect the original image, and obtain the image features of the original image based on the image feature analysis algorithm;

[0054] Step 2: Perform semantic analysis on the image features, and identify and extract ethnic cultural symbol information;

[0055] Step 3: Generate directional index score values of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree according to the ethnic cultural symbol information;

[0056] Step 4: Integrate all the directional index score values to generate the ethnic harmony index of the original image;

[0057] Step 5: Based on the visualization technology, generate the ethnic communication evaluation result in combination with the ethnic harmony index and the directional index score values of the original image.

[0058] In this embodiment, the original image refers to the initial image file used for analysis and evaluation in this technical solution, and is an image containing ethnic cultural symbols, elements, graphics, etc. that need to be recognized and analyzed, including: photos, paintings, icons, design works, etc.

[0059] In this embodiment, the image feature analysis algorithm is a computational method for extracting and analyzing key information in an image. By identifying various features in the image, such as color, texture, shape, edge, corner points, etc., the image is converted into a digital feature vector.

[0060] In this embodiment, the image feature refers to the representative information or attributes extracted from the image.

[0061] In this embodiment, semantic analysis is a process of understanding and interpreting information, aiming to extract hidden and meaningful content from the original image.

[0062] In this embodiment, the ethnic cultural symbol information refers to various recognizable symbols, elements, features or signs associated with a specific ethnic group or cultural community.

[0063] In this embodiment, the direction index is an index system used to quantify and evaluate different aspects related to ethnic harmony, including: spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree and psychological identity degree.

[0064] In this embodiment, the ethnic harmony index is a quantitative index used to evaluate the degree of interaction, integration and harmonious coexistence of different ethnic cultures, groups or social members in the image.

[0065] In this embodiment, the visualization technology refers to the conversion of data, information, analysis results, etc. into an intuitive and easy-to-understand visual form through visual means such as graphics, images, animations, etc.

[0066] In this embodiment, the ethnic communication evaluation result is a visual representation that comprehensively evaluates the integration degree of different ethnic cultures in the image, the social harmony situation and their cultural identity based on the ethnic harmony index and the direction index score values of the original image.

[0067] The working principle and beneficial effects of the above technical solution are: through image feature extraction and semantic analysis, ethnic cultural symbols are identified, multiple social and cultural indicators are generated and integrated into the ethnic harmony index, and finally the evaluation result is presented visually, which helps to promote cultural integration, social harmony and in-depth understanding and promotion of ethnic communication.

[0068] Embodiment 2:

[0069] On the basis of the above Embodiment 1, after collecting the original image, it includes:

[0070] According to the preset target data source, collect the original image as the analysis target;

[0071] Perform gray-scale processing on each pixel point of the original image to generate the first image:

[0072] Use a Gaussian filter kernel to perform smoothing processing on the first image, take the weighted average of the pixels in its neighborhood for each pixel point, and then perform smoothing processing on the pixel point according to the weighted average combined with the Gaussian filter kernel to obtain the second image after filtering to remove noise.

[0073] In this embodiment, the target data source refers to the data source for obtaining the original image, such as photos, videos, portraits, etc.

[0074] In this embodiment, grayscale processing is a common operation in image processing. Its purpose is to convert a color image into a grayscale image. Specifically: ; where represents the grayscale value of the pixel point (x, y), represents the preset sensitive red coefficient, represents the preset sensitive green coefficient, represents the preset sensitive blue coefficient, represents the red component of the pixel point (x, y), represents the green component of the pixel point (x, y), represents the blue component of the pixel point (x, y).

[0075] In this embodiment, the first image is the original image after grayscale processing.

[0076] In this embodiment, the grayscale value is the brightness value of each pixel in the image, indicating the light and dark degree of the pixel.

[0077] In this embodiment, the preset sensitive red coefficient is a coefficient used to weight the red component when calculating the grayscale value of the image.

[0078] In this embodiment, the preset sensitive green coefficient is a coefficient used to weight the green component when calculating the grayscale value of the image.

[0079] In this embodiment, the preset sensitive blue coefficient is a coefficient used to weight the blue component when calculating the grayscale value of the image.

[0080] In this embodiment, the red component refers to the red color information of each pixel point in the image.

[0081] In this embodiment, the green component refers to the green color information of each pixel point in the image.

[0082] In this embodiment, the blue component refers to the blue color information of each pixel point in the image.

[0083] In this embodiment, the Gaussian filter kernel is a tool for image smoothing processing and is applied in image processing for denoising, smoothing, or blurring the image.

[0084] In this embodiment, smoothing processing refers to processing the image through filtering technology to remove noise and smooth the details in the image, making the image look softer and the details more coherent. The process of smoothing the specific pixel points is as follows: ; where represents the pixel value after smoothing processing of the pixel point (x, y), represents the neighborhood pixel values of the pixel point (x, y) based on the i and j offsets, K represents the width of the first image, and H represents the height of the first image, represents the Gaussian filter kernel, represents the standard deviation of the first image, represents the exponential function, represents the pi.

[0085] In this embodiment, the second image is an image with noise removed and details smoother after Gaussian filtering processing.

[0086] The working principle and beneficial effects of the above technical solution are: by performing grayscale processing on the original image to generate the first image; then using the Gaussian filter kernel for smoothing processing to remove noise and generate the second image. This process effectively improves the image quality and provides a clear data basis for subsequent analysis.

[0087] Embodiment 3:

[0088] Based on the above Embodiment 2, image features of the original image are obtained based on an image feature analysis algorithm, including:

[0089] Construct a horizontal direction operator and a vertical direction operator according to a preset template, and calculate the gradient magnitude of any pixel of the second image based on the horizontal direction operator and the vertical direction operator;

[0090] Take any pixel point of the second image as the central pixel point, obtain the surrounding pixel points of the central pixel point according to a preset distance, and calculate the neighborhood pixel gradient magnitude difference matrix between the central pixel point and any surrounding pixel point, and statistically analyze the neighborhood pixel gradient magnitude difference matrices of all pixel points;

[0091] Input the neighborhood pixel gradient magnitude difference matrix of any pixel point into a preset response function and output a response value. If the response value is greater than a preset threshold, determine that the pixel point is an edge point;

[0092] All edge points of the second image constitute the image features of the second image.

[0093] In this embodiment, the preset template refers to a set of preset filters or operators for specific operations.

[0094] In this embodiment, the horizontal direction operator is an operator used to detect horizontal edges in the image and measure the change in the horizontal direction of the image.

[0095] In this embodiment, the vertical direction operator is an operator used to detect vertical edges in the image and measure the change in the vertical direction of the image.

[0096] In this embodiment, the pixel gradient magnitude is a measure representing the intensity change around a certain pixel point in the image. The specific acquisition method is as follows: ; where represents the gradient magnitude of the second image pixel point (x, y), represents the pixel value of the second image pixel point (x, y), represents the horizontal direction operator, represents the vertical direction operator.

[0097] In this embodiment, the preset distance refers to the range preset in calculating the neighborhood of pixels.

[0098] In this embodiment, the neighborhood pixel gradient magnitude difference matrix is a matrix used to measure the gradient difference of neighborhood pixels around a certain pixel point in the image.

[0099] In this embodiment, the response function is a mathematical expression used to evaluate whether a pixel point is an edge point, and generates a response value corresponding to the center point by inputting the neighborhood pixel gradient magnitude difference matrix of the center point.

[0100] In this embodiment, an edge point refers to a pixel point with significant brightness change in the image, representing the division of the boundary region of an object in the image.

[0101] The working principle and beneficial effects of the above technical solution are as follows: By calculating the gradient magnitude of image pixels and judging edge points through the neighborhood pixel gradient difference matrix, image features are extracted. This method can effectively identify the edge information in the image, which helps to improve the accuracy of image processing and analysis.

[0102] Embodiment 4:

[0103] On the basis of the above Embodiment 1, semantic analysis is performed on the image features to identify and extract ethnic cultural symbol information, including:

[0104] Obtain the expert analysis results of ethnic cultural symbols and construct templates of multiple cultural symbol semantic types;

[0105] Slide and match any template on the image features, and generate a sliding match similarity each time it slides;

[0106] Obtain the maximum sliding match similarity between the template and the image features according to the preset maximum value function. If the maximum sliding match similarity is greater than the preset standard match threshold, it is determined that the image features have the cultural symbol semantic type corresponding to the template, otherwise it means it does not exist;

[0107] Based on the matching results of the described image features and all templates, all the matching ethnic cultural symbol semantic types and the maximum sliding matching similarity are obtained, constituting the ethnic cultural symbol information of the image features.

[0108] In this embodiment, an ethnic cultural symbol is a unique symbolic mark, image, symbol, or symbol system that a nation or cultural group has in aspects such as its history, tradition, custom, language, religion, etc.

[0109] In this embodiment, the expert analysis results include in-depth research on ethnic cultural symbols and summaries of their symbolic meanings. For example, certain symbols may represent religious beliefs, historical events, natural scenes, etc.

[0110] In this embodiment, the semantic type of a cultural symbol refers to the specific meaning or symbol conveyed by the cultural symbol, including: a certain specific concept, belief, custom, or value under a certain cultural group, historical, religious, or social background.

[0111] In this embodiment, sliding matching is performed by gradually comparing a template with the target image, calculating the similarity between the template and the image features at different positions, and finding the best matching position.

[0112] In this embodiment, the sliding matching similarity is an index to measure the similarity between the template and the image features during the sliding matching process. The specific acquisition method is as follows: ; where represents the sliding matching similarity between the template and the image features, represents the matching degree function, represents the pixel value of the pixel point I(x, y) in the image features and the corresponding pixel point of the template. p represents the height of the template, and q represents the width of the template. represents the sliding window system of the size of the template. k represents the abscissa of the corresponding pixel point between the template and the image features, and q represents the ordinate of the corresponding pixel point between the template and the image features.

[0113] In this embodiment, the preset maximum value function refers to a function used to find the maximum matching similarity value from all sliding positions. For example ; represents based on the module obtained maximum sliding matching similarity.

[0114] In this embodiment, the preset standard matching threshold is a parameter used to judge whether the matching result between the image features and the template is significant during the cultural symbol matching process.

[0115] The working principle and beneficial effects of the above technical solution are as follows: By sliding to match multiple cultural symbol templates with image features, calculating the similarity, and determining whether there is a cultural symbol semantic type, the ethnic cultural symbol information is finally extracted. This method enhances the recognition and analysis accuracy of cultural symbols in images and helps with the extraction of cultural features.

[0116] Example 5:

[0117] Based on the above Example 4, the determination of the indicators of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree includes:

[0118] According to all cultural symbol semantic types, corresponding secondary indicators are generated;

[0119] Using a word embedding model, input all secondary indicators and five preset direction indicators, where the direction indicators include: spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree, to generate corresponding semantic vector values;

[0120] Taking any one of the direction indicators as the center for clustering, obtaining the semantic similarity between any one secondary indicator and all center clusters, and determining the center cluster with the highest semantic similarity as the direction indicator to which the secondary indicator belongs;

[0121] According to all secondary indicators corresponding to any one direction indicator, determining the semantic similarity between any two secondary indicators, those with a similarity greater than the first threshold are classified into the same unit class, generating multiple unit classes under the direction indicator, and naming any one unit class according to the semantic characteristics of the unit class to obtain the first-level indicator;

[0122] Through the social identity theory, a questionnaire survey is conducted on a preset group to obtain the voting evaluations of the first-level indicators, secondary indicators, and direction indicators. Combining data statistics methods to quantify the voting evaluations to obtain the weights of the first-level indicators, secondary indicators, and direction indicators, and obtaining the standard values of the secondary indicators from national data sources.

[0123] In this example, the secondary indicators are the specific refinement and expansion of the first-level indicators in a multi-dimensional and hierarchical analysis framework. For example: the proportion of the floating population of ethnic minorities, the ethnic minority distribution index, the degree of use of social public spaces, etc.

[0124] In this example, the word embedding model is a natural language processing technology used to convert words or phrases into low-dimensional vectors in a vector space to capture the semantic relationships between words.

[0125] In this example, the semantic vector value is the vector representation of words, sentences, or texts converted into a high-dimensional space through a word embedding model.

[0126] In this embodiment, central clustering refers to a representative index or semantic vector used to represent a clustering group in cluster analysis.

[0127] In this embodiment, semantic similarity refers to the degree of semantic similarity between two or more words, sentences, or texts. The specific acquisition method is as follows: ; where CCD represents the semantic similarity between the secondary index and the central clustering, represents the clustering function, represents the semantic vector value of the secondary index, represents the semantic vector value of the central clustering.

[0128] In this embodiment, the clustering function is used to measure the similarity degree of an object or element in a specific space. For example: ; where represents the clustering function of vectors a and b, represents the dot product, represents the modulus of vector a, represents the modulus of vector b.

[0129] In this embodiment, the unit class refers to a group of secondary indicators that are semantically similar after classification based on semantic similarity.

[0130] In this embodiment, the primary indicator refers to an indicator at the macro and basic level used to measure and evaluate a specific social, cultural, or economic phenomenon. For example: the index of inter-embeddedness of the populations of all ethnic groups, the employment and insurance participation rate of ethnic minorities, the degree of understanding of Chinese culture by ethnic minorities, etc.

[0131] In this embodiment, social identity theory is a theory about group behavior, interaction among group members, and how individuals define their identities based on the social groups they belong to.

[0132] In this embodiment, the data statistical method describes the distribution of data through data statistical characteristics (such as mean, variance, correlation coefficient, etc.).

[0133] In this embodiment, the standard value of the secondary indicator refers to the quantified value of the secondary indicator in a specific evaluation system, which reflects the standardization degree of the indicator.

[0134] In this embodiment, national data sources include: data from the statistical bureau, statistical data of government departments such as education, culture, and sports, data in fields such as social welfare, medical security, housing, and public safety, etc.

[0135] The working principle and beneficial effects of the above technical solution are as follows: By generating semantic vectors through a word embedding model, calculating the similarity between secondary indicators and direction indicators, performing clustering and classification, combining social identity theory and data statistics methods to quantify the weights of each indicator, and setting standard values based on national data, the accuracy and systematicness of indicator evaluation and cultural analysis are improved.

[0136] Example 6:

[0137] Based on the above Example 5, generate score values for spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree, including:

[0138] Based on the ethnic cultural symbol information of the image features, calculate the score value of any first-level indicator of the image features;

[0139] Based on the score values of all first-level indicators and their weights, generate the score values of the direction indicators for spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree, and psychological recognition degree.

[0140] In this example, the way to obtain the score value of any first-level indicator is: ; where, represents the score value of any first-level indicator of the image features, N represents the total number of secondary indicators in the first-level indicators, represents the weight of the i1-th secondary indicator, represents the sliding matching similarity of the i1-th secondary indicator, represents the standard value of the i1-th secondary indicator, represents the score value of the i1-th secondary indicator.

[0141] The working principle and beneficial effects of the above technical solution are as follows: By calculating the score of the first-level indicator of the image features and combining the weight and sliding matching similarity of the secondary indicators, generate scores in the directions of spatial inter-embedding degree, economic co-prosperity degree, etc., providing accurate ethnic cultural symbol analysis and multi-dimensional evaluation, and enhancing the systematicness and accuracy of cultural research.

[0142] Example 7:

[0143] Based on the above Example 6, fuse all the score values of the direction indicators to generate the ethnic harmony index of the original image, including:

[0144] Based on the five direction indicators and their weights, perform weighted fusion calculation according to the score values to generate the ethnic harmony index of the original image.

[0145] In this example, the weighted fusion calculation is obtained based on the five direction indicators and their weights, specifically: ; where RRT represents the ethnic harmony index, represents the weight coefficient of the spatial inter-embedding degree, represents the fractional value of the spatial inter-embedding degree, represents the weight coefficient of the economic co-prosperity degree, represents the fractional value of the economic co-prosperity degree, represents the weight coefficient of the cultural integration degree, represents the fractional value of the cultural integration degree, represents the weight coefficient of the social harmony degree, represents the fractional value of the social harmony degree, represents the weight coefficient of the psychological recognition degree, represents the fractional value of the psychological recognition degree;

[0146] The working principle and beneficial effects of the above technical solution are: By integrating the five-direction indicators and their weights, calculating the ethnic harmony index based on the fractional values helps to quantify the cultural symbol performance of the original image, provides a comprehensive and accurate ethnic culture assessment, and is beneficial to cultural diversity research and the construction of a harmonious society.

[0147] Example 8:

[0148] Based on the above Example 1, based on visualization technology, generate the ethnic communication assessment results of the original image, including:

[0149] Based on visualization technology, take the ethnic harmony index of the original image and the fractional values of all direction indicators as the main display part of the visualization report; take any direction indicator and the corresponding index value as the subordinate display part, select any direction indicator dropdown to jointly display all first-level indicators - second-level indicators, and construct a multi-level visualization intelligent recognition and analysis page for cultural symbols;

[0150] Based on the blank report template of the multi-level visualization intelligent recognition and analysis page for cultural symbols, input the second-level indicators and their index values of the original image, and output the ethnic communication assessment results of the original image.

[0151] In this example, the main display part refers to the most core and important part in the entire report page, carrying the most critical data and results of image analysis.

[0152] In this example, the subordinate display part refers to the additional and detailed information level provided in the visualization report, in addition to displaying the core information of the main ethnic harmony index and all direction indicators.

[0153] In this example, the report template provides a standardized way for users to input and display the analysis data of the original image.

[0154] The working principle and beneficial effects of the above technical solution are as follows: The visualization technology is used to display the ethnic harmony index and the scores of various directional indicators. By multi-level display of the first-level and second-level indicators, intuitive cultural symbol analysis and intelligent recognition are provided. Combining with the report template, the ethnic communication evaluation results are output, improving the accuracy of cultural analysis and the user experience.

[0155] Example 9:

[0156] As Figure 2 shown, the cultural symbol intelligent recognition and analysis system based on image semantic understanding provided by the embodiment of the present invention includes:

[0157] Image feature module: Collect the original image and obtain the image features of the original image based on the image feature analysis algorithm;

[0158] Extraction module: Perform semantic analysis on the image features, identify and extract ethnic cultural symbol information;

[0159] Direction index module: Generate the directional index score values of spatial inter-embedding degree, economic co-prosperity degree, cultural integration degree, social harmony degree and psychological identity degree according to the ethnic cultural symbol information;

[0160] Ethnic harmony index module: Integrate all the directional index score values to generate the ethnic harmony index of the original image;

[0161] Visualization module: Based on the visualization technology, combine the ethnic harmony index and the directional index score values of the original image to generate the ethnic communication evaluation results.

[0162] The working principle and beneficial effects of the above technical solution are as follows: By extracting image features and semantic analysis, ethnic cultural symbols are identified, multiple social and cultural indicators are generated and integrated into the ethnic harmony index, and finally the evaluation results are presented visually, which helps to promote cultural integration, social harmony and in-depth understanding and promotion of ethnic communication.

[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. The method of intelligent recognition and analysis of cultural symbols based on image semantic understanding is characterized by: include: Step 1: Collect the original image and obtain the image features of the original image based on the image feature analysis algorithm; Step 2: Perform semantic analysis on image features to identify and extract national cultural symbol information; Step 3: Generate directional index scores of spatial embeddedness, economic common prosperity, cultural integration, social harmony and psychological identification based on the information of ethnic cultural symbols; Step 4: Fuse all the directional index scores to generate the ethnic harmony index of the original image; Step 5: Based on visualization technology, the ethnic harmony index and direction index score of the original image are combined to generate the ethnic communication evaluation results.

2. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 1 is characterized in that: After collecting the original image, including: According to the preset target data source, the original image is collected as the analysis target; Perform grayscale processing on each pixel of the original image to generate the first image: The first image is smoothed using a Gaussian filter kernel, and the weighted average of the pixels in its neighborhood is taken for each pixel. The pixel is then smoothed based on the weighted average combined with the Gaussian filter kernel to obtain a second image after filtering and removing noise.

3. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 2 is characterized in that: The image features of the original image are obtained based on the image feature analysis algorithm, including: Constructing a horizontal direction operator and a vertical direction operator according to a preset template, and calculating a gradient amplitude of any pixel of the second image based on the horizontal direction operator and the vertical direction operator; Taking any pixel point of the second image as the central pixel point, obtaining the pixel points around the central pixel point according to the preset distance, and calculating the neighborhood pixel gradient amplitude difference matrix of the central pixel point and any surrounding pixel point, and counting the neighborhood pixel gradient amplitude difference matrix of all pixel points; Inputting the neighborhood pixel gradient amplitude difference matrix of any pixel point into a preset response function and outputting a response value, if the response value is greater than a preset threshold, determining the pixel point as an edge point; All edge points of the second image constitute image features of the second image.

4. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 1, characterized in that: Perform semantic analysis on image features to identify and extract national cultural symbol information, including: Obtain expert analysis results of national cultural symbols and construct templates of multiple cultural symbol semantic types; Perform sliding matching on any template on the image features, and generate sliding matching similarity each time; Obtaining the maximum sliding matching similarity between the template and the image feature according to a preset maximum value function, and if the maximum sliding matching similarity is greater than a preset standard matching threshold, determining that the image feature has the cultural symbol semantic type corresponding to the template, otherwise it indicates that the cultural symbol semantic type does not exist; According to the matching results of the image feature and all templates, the semantic types of all matched cultural symbols and the maximum sliding matching similarity are obtained to constitute the national cultural symbol information of the image feature.

5. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 4 is characterized in that: The lower-level indicators of spatial embedding, economic common prosperity, cultural integration, social harmony and psychological identity are determined, including: Generate corresponding secondary indicators based on all cultural symbol semantic types; Using the word embedding model, all secondary indicators and five preset directional indicators are input, including spatial embedding, economic co-enrichment, cultural integration, social harmony and psychological identification, to generate corresponding semantic vector values; Taking any directional indicator as the central cluster, obtaining the semantic similarity between any secondary indicator and all central clusters, and determining the central cluster with the highest semantic similarity as the directional indicator to which the secondary indicator belongs; According to all the secondary indicators corresponding to any directional indicator, the semantic similarity of any two secondary indicators is determined, and those greater than the first threshold are divided into the same unit class, multiple unit classes under the directional indicator are generated, and any unit class is named according to the semantic features of the unit class to obtain the primary indicator; Through social identity theory, a questionnaire survey is conducted on the preset groups to obtain the voting evaluation of the primary indicators, secondary indicators and directional indicators. The voting evaluation is quantified by combining data statistical methods to obtain the weights of the primary indicators, secondary indicators and directional indicators, and the standard values ​​of the secondary indicators are obtained from the national data source.

6. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 5 is characterized in that: Generate directional indicator scores for spatial embeddedness, economic common prosperity, cultural integration, social harmony, and psychological identity, including: Based on the national cultural symbol information of the image feature, calculating the score value of any first-level index of the image feature; Based on the scores and weights of all primary indicators, directional indicator scores of spatial embeddedness, economic common prosperity, cultural integration, social harmony and psychological identification are generated.

7. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 6, characterized in that: All directional index scores are combined to generate the ethnic harmony index of the original image, including: Based on the five directional indicators and their weights, a weighted fusion calculation is performed according to the directional indicator score values ​​to generate the ethnic harmony index of the original image.

8. The method for intelligent identification and analysis of cultural symbols based on image semantic understanding according to claim 1, characterized in that: Based on visualization technology, the ethnic communication assessment results of the original image are generated, including: Based on visualization technology, the ethnic harmony index of the original image and the score values ​​of all directional indicators are used as the main display part of the visualization report; any directional indicator and the corresponding score value are used as the subordinate display part, and all first-level indicators and second-level indicators are jointly displayed by selecting any directional indicator and pulling down, thus constructing a multi-level visualized cultural symbol intelligent recognition and analysis page; A blank report template for the intelligent identification and analysis page of cultural symbols based on multi-level visualization, inputs the secondary indicators and their score values ​​of the original image, and outputs the ethnic communication assessment results of the original image.

9. A system for intelligent recognition and analysis of cultural symbols based on image semantic understanding, applied to a method for intelligent recognition and analysis of cultural symbols based on image semantic understanding as claimed in any one of claims 1 to 8, characterized in that: include: Image feature module: collects original images and obtains image features of the original images based on image feature analysis algorithms; Extraction module: performs semantic analysis on image features, identifies and extracts national cultural symbol information; Direction indicator module: Generates the direction indicator scores of spatial embedding, economic common prosperity, cultural integration, social harmony and psychological identification based on the information of national cultural symbols; National harmony index module: merges all directional index scores to generate the national harmony index of the original image; Visualization module: Based on visualization technology, the ethnic harmony index and direction indicator score values ​​of the original image are combined to generate ethnic communication evaluation results.

Images