A tree age identification system and identification method based on image recognition

Abstract

The present invention relates to the technical field of tree protection, and specifically to a tree age identification system and identification method based on image recognition. The system includes an image texture analysis and identification module, an environmental influencing factor analysis module, a spectral feature extraction and analysis module, a tree type analysis and identification module, a climate feature impact analysis module, and a tree age prediction model construction module. In the present invention, by analyzing the annual ring image data and identifying the texture features, the reliability of tree age estimation is improved, the influence of environmental factors on the annual ring shape is combined to improve the adaptability and accuracy of the prediction, the spectral response of the trees is collected and analyzed, the ability to classify tree species is improved, and the spectral vector feature analysis is combined to effectively distinguish a variety of tree species. By evaluating the influence of climatic conditions on the width and density of annual rings, the adaptability of the model to environmental changes is enhanced. The tree age dynamic prediction model constructed in combination with geographical and climatic factors improves the accuracy and reliability of tree age estimation.

Description

Technical Field

[0001] The present invention relates to the technical field of tree protection, and in particular to a tree age identification system and an identification method based on image recognition. Background Art

[0002] The field of tree protection technology includes the assessment of tree health, prevention and control of diseases, and optimization of tree growth environment. It uses physical and chemical methods and information technology, including remote sensing, biotechnology, and computer-assisted image processing, to maintain and restore forest ecosystems, effectively monitor the health status of trees, promptly discover and deal with problems, and promote the sustainable management and utilization of tree resources. It is of great significance to maintaining ecological balance, maintaining biodiversity, and promoting ecological sustainable development.

[0003] Among them, the tree age identification system based on image recognition estimates and analyzes the age of trees by utilizing image processing technology, including the use of automated image recognition methods to reduce reliance on traditional manual measurement methods, thereby improving the efficiency and accuracy of tree age identification. By automatically identifying tree rings and analyzing the annual ring characteristics of a variety of tree species, the rapid and accurate determination of tree age can be achieved. This plays a key role in forestry management, ecological research, and environmental monitoring, helping to scientifically manage tree resources and support ecological protection and research work.

[0004] Traditional tree age identification technology relies on physical and chemical methods and information technology, including remote sensing, biotechnology and computer-assisted image processing. It still has limitations in the accuracy and environmental adaptability of tree age identification. It relies on manual measurement and simple automated methods, cannot fully utilize environmental and spectral data, limits the accuracy of prediction models and their ability to adapt to environmental changes, cannot distinguish the impact of climate change on tree ring characteristics, and cannot accurately reflect the actual impact of climate factors on tree growth patterns, resulting in deviations between tree age estimation results and actual conditions, affecting the quality of decision-making in forestry management and ecological research, and having a negative impact on ecological protection and resource management. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a tree age identification system and identification method based on image recognition.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme: a tree age identification system based on image recognition comprises:

[0007] The image texture analysis and recognition module analyzes the texture features in tree ring images based on tree ring image data, including differences in cell structures, and generates a tree texture feature set;

[0008] The environmental impact factor analysis module analyzes the influence of humidity, temperature, and light on the shape of annual rings based on the tree texture feature set, and generates an environmental response feature analysis result;

[0009] The spectral feature extraction and analysis module collects the spectral response information of trees based on the environmental response feature analysis result, analyzes the response characteristics of various trees, and generates tree spectral classification vector information;

[0010] The tree type analysis and identification module uses the tree spectral classification vector information to identify tree species characteristics and classify tree species by analyzing the characteristic intensities and distribution patterns of multiple spectral vectors, thereby generating a tree species identification feature information set;

[0011] The climate characteristic impact analysis module evaluates the growth pattern of tree rings under differentiated climate conditions based on the tree species identification characteristic information set, analyzes the impact of climate change on tree ring width and density, and generates tree ring growth pattern analysis results;

[0012] The tree age prediction model construction module constructs a tree age dynamic prediction model based on the analysis results of the annual ring growth pattern and combines geographical and climatic factors to generate tree age identification information by combining the basic information of the annual rings identified by the image texture analysis and recognition module, including the number of annual rings and effective sample length.

[0013] As a further scheme of the present invention, the tree texture feature set includes cell wall thickness information, cell arrangement pattern information, and cell size variation information; the environmental response feature analysis results include analysis results of the impact of light cycle on annual ring shape, analysis results of the impact of humidity changes on annual ring density, and analysis results of the impact of temperature fluctuations on annual ring color; the tree spectral classification vector information specifically includes tree species spectral absorption peak information, tree species spectral reflection peak information, and tree species spectral intensity distribution information; the tree species identification feature information set includes tree species spectral fingerprint data, tree species morphological characteristics, and tree species ecological preferences; the annual ring growth pattern analysis results include the impact of temperature increase on annual ring thickness, the impact of annual average precipitation on annual ring density, and the impact of seasonal temperature difference on annual ring growth rate; the tree age identification information specifically includes tree age range prediction information, growth start year prediction information, and tree age prediction error range information.

[0014] As a further solution of the present invention, the image texture analysis and recognition module includes:

[0015] The texture feature recording submodule analyzes the texture details of multiple tree ring images based on tree ring image data, identifies the continuity, fracture pattern and color information of the texture, and generates tree ring texture analysis information;

[0016] The cell feature extraction submodule analyzes the cell structure differences in the tree ring image based on the tree ring pattern analysis information, including cell size, shape and mutual arrangement pattern, and generates tree cell identification data;

[0017] The texture trend analysis submodule analyzes the annual ring texture features of various trees with different tree ages based on the tree cell identification data, identifies texture change trends and features, and generates a tree texture feature set.

[0018] As a further solution of the present invention, the environmental impact factor analysis module includes:

[0019] The environmental data collection submodule collects environmental data associated with tree growth based on the tree texture feature set, including average temperature and humidity and light intensity of the area, and generates tree area environmental analysis results;

[0020] The influencing factor analysis submodule evaluates the impact of temperature changes and light intensity on tree growth based on the tree area environmental analysis results, and generates environmental interaction evaluation data;

[0021] The environmental characteristic integration submodule analyzes the impact of various environmental factors on annual ring characteristics based on the environmental interaction assessment data, identifies the relationship with tree density and annual ring width, and generates environmental response characteristic analysis results.

[0022] As a further solution of the present invention, the spectral feature extraction and analysis module includes:

[0023] The spectral data acquisition submodule performs spectral scanning of trees at multiple wavelengths based on the environmental response characteristic analysis results, analyzes the reflection and absorption data of the multiple wavelengths, and generates a multi-spectral scanning data set;

[0024] The spectral response analysis submodule analyzes the intensity of spectral reflection and absorption of multiple bands based on the multi-spectral scanning data set, identifies the characteristic points of spectral data, and generates spectral information analysis results;

[0025] The spectral feature encoding submodule converts the analyzed spectral features, including spectral peaks, absorption bands and reflectance, into digital vectors based on the spectral information analysis results, analyzes the spectral response characteristics of various tree types, and generates tree spectral classification vector information.

[0026] As a further solution of the present invention, the tree type analysis and identification module includes:

[0027] The tree species spectrum comparison submodule uses the tree spectrum classification vector information to match with the historical tree species spectrum data, evaluates the matching degree between the spectrum vectors of multiple trees and the historical data, and generates spectrum data matching results;

[0028] The tree species characteristic identification submodule analyzes the spectral peaks and key wavelength ranges of various trees through the spectral data matching results, identifies the spectral fingerprints of various trees and records the tree species characteristics, and generates tree species characteristic analysis information;

[0029] The tree species classification result integration submodule constructs a tree species feature database based on the tree species characteristic analysis information, including the tree's spectral fingerprint, annual ring texture characteristics and annual ring density change characteristics, and generates a tree species identification feature information set.

[0030] As a further solution of the present invention, the climate characteristics impact analysis module includes:

[0031] The climate data analysis submodule analyzes historical climate data associated with tree growth based on the tree species identification feature information set, including temperature change patterns, humidity change patterns, and annual precipitation, analyzes the periodic change patterns of climate conditions, and generates historical meteorological information analysis results;

[0032] The tree ring impact assessment submodule analyzes the relationship between the climate data of multiple years and the change of tree ring width and density based on the historical meteorological information analysis results, identifies the impact of climate factors on tree ring growth, and generates climate factor impact assessment information;

[0033] The growth factor assessment submodule analyzes the impact of climate change on the growth patterns of various tree rings, including tree ring width, tree density, and tree ring color, based on the climate factor impact assessment information, and generates a growth ring growth pattern analysis result.

[0034] As a further solution of the present invention, the tree age prediction model construction module includes:

[0035] The tree age data preparation submodule analyzes the historical growth records of trees and corresponding climate data, including growth ring width, tree density, tree height, precipitation, and temperature changes, based on the analysis results of the tree ring growth pattern, and generates a tree age modeling data set;

[0036] The growth factor analysis submodule analyzes the influence of climate conditions on tree ring formation based on the tree age modeling data set, identifies key climate factors that affect tree ring growth, and generates climate influencing factor identification information;

[0037] The age model construction submodule uses a random forest algorithm based on the climate influencing factor identification information to construct a tree age prediction model, uses tree ring images and environmental factors to predict tree age, and generates tree age identification information.

[0038] As a further solution of the present invention, the random forest algorithm is according to the formula:

[0039] ;

[0040] Calculate the tree age, where is the index used to traverse all decision tree models, is the total number of decision trees, For the The weight of the decision tree model, is the input tree ring image data, To predict tree age, For the A decision tree takes into account environmental parameters , , The prediction results after is the environmental stability parameter, is the light intensity, is the coefficient of variation of precipitation.

[0041] A tree age identification method based on image recognition, the tree age identification method based on image recognition is performed based on the above-mentioned tree age identification system based on image recognition, and comprises the following steps:

[0042] S1: Based on tree ring image data, analyze the texture characteristics of tree rings, including cell structure differences and texture patterns, and generate tree ring texture analysis information;

[0043] S2: using the tree ring texture analysis information, analyzing the effects of various external environments on tree ring shape and growth, including precipitation and light intensity, to generate environmental regulation texture impact data;

[0044] S3: Based on the environmental adjustment texture impact data, the spectral reflectance and absorptivity of various trees are collected, the spectral response analysis of the trees is performed, the spectral characteristics of the tree species are identified, and a spectral response feature set is generated;

[0045] S4: using the spectral response feature set to compare with a spectral database of known tree species, matching and classifying tree species, and generating a tree species feature classification result;

[0046] S5: Based on the classification results of tree species characteristics, evaluate the long-term impact of climate change on the growth patterns of tree rings of various trees, build a dynamic tree age prediction model, and combine basic tree ring information, including the number of tree rings and effective sample length, to generate tree age identification information.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are:

[0048] In the present invention, by analyzing the annual ring image data and identifying texture features, the understanding and utilization of the differences in tree cell structures are enhanced, the reliability of tree age estimation is improved, the influence of environmental factors on the shape of annual rings is integrated, the adaptability and accuracy of prediction are improved, the spectral response of trees is collected and analyzed, the ability to classify tree species is enhanced, and spectral vector feature analysis is combined to effectively distinguish multiple tree species. By evaluating the influence of climatic conditions on the width and density of annual rings, the tree age prediction model can take into account the long-term effects of climate change and enhance the model's adaptability to environmental changes. The dynamic tree age prediction model constructed by combining geographical and climatic factors improves the accuracy and reliability of tree age estimation and optimizes the comprehensiveness and practicality of the prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a system flow chart of the present invention;

[0050] Figure 2 It is a schematic diagram of the system framework of the present invention;

[0051] Figure 3 This is a flow chart of the image texture analysis and recognition module of the present invention;

[0052] Figure 4 It is a flow chart of the environmental impact factor analysis module of the present invention;

[0053] Figure 5 It is a flow chart of the spectrum feature extraction and analysis module of the present invention;

[0054] Figure 6 This is a flow chart of the tree type analysis and identification module of the present invention;

[0055] Figure 7 This is a flow chart of the climate characteristics impact analysis module of the present invention;

[0056] Figure 8 A flowchart of a tree age prediction model construction module of the present invention;

[0057] Fig. 9 It is a schematic diagram of the method steps of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0059] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0060] Embodiment 1

[0061] See also Figure 1 to Figure 2 The present invention provides a technical solution, a tree age identification system based on image recognition, comprising:

[0062] The image texture analysis and recognition module analyzes the texture features in tree ring images based on tree ring image data, including differences in cell structures, and generates a tree texture feature set;

[0063] The environmental impact factor analysis module analyzes the effects of humidity, temperature, and light on the shape of tree rings based on the tree texture feature set, and generates environmental response feature analysis results;

[0064] The spectral feature extraction and analysis module collects the spectral response information of trees based on the environmental response feature analysis results, analyzes the response characteristics of various trees, and generates tree spectral classification vector information;

[0065] The tree type analysis and identification module uses tree spectral classification vector information to identify tree species characteristics and classify tree species by analyzing the characteristic intensity and distribution patterns of multiple spectral vectors, and generates a tree species identification feature information set;

[0066] The climate characteristics impact analysis module evaluates the growth pattern of tree rings under differentiated climate conditions based on the tree species identification feature information set, analyzes the impact of climate change on tree ring width and density, and generates tree ring growth pattern analysis results;

[0067] The tree age prediction model construction module builds a dynamic tree age prediction model based on the analysis results of the annual ring growth pattern and combines geographical and climatic factors to generate tree age identification information by combining the basic annual ring information identified by the image texture analysis and recognition module, including the number of annual rings and effective sample length.

[0068] The tree texture feature set includes cell wall thickness information, cell arrangement pattern information, and cell size variation information. The environmental response feature analysis results include the analysis results of the impact of light cycle on annual ring shape, the analysis results of the impact of humidity changes on annual ring density, and the analysis results of the impact of temperature fluctuations on annual ring color. The tree spectral classification vector information specifically includes tree species spectral absorption peak information, tree species spectral reflection peak information, and tree species spectral intensity distribution information. The tree species identification feature information set includes tree species spectral fingerprint data, tree species morphological characteristics, and tree species ecological preferences. The annual ring growth pattern analysis results include the impact of rising temperature on annual ring thickness, the impact of annual average precipitation on annual ring density, and the impact of seasonal temperature difference on annual ring growth rate. The tree age identification information specifically includes tree age range prediction information, growth start year prediction information, and tree age prediction error range information.

[0069] See also Figure 2 and Figure 3 , the image texture analysis and recognition module includes:

[0070] The texture feature recording submodule analyzes the texture details of multiple tree ring images based on the tree ring image data, identifies the continuity, fracture pattern and color information of the texture, and generates the specific process of tree ring texture analysis information as follows;

[0071] In the texture feature recording submodule, based on the annual ring image data, by collecting the annual ring image data of differentiated tree species, each image is scanned at high resolution, and the image processing algorithm is used to analyze the texture details, including the width, depth and continuity of the texture. The image is grayed, and the edge detection technology is used to identify the fracture pattern. The color analysis algorithm is used to quantify the color change of the annual ring. The texture continuity evaluation model is adopted, and the formula is: ,in, represents the texture continuity index, For the The weight of the pixel, is the texture feature value of the corresponding pixel point, and the annual ring texture analysis information is obtained.

[0072] The cell feature extraction submodule analyzes the differences in cell structure in the tree ring image based on the tree ring pattern analysis information, including cell size, shape and mutual arrangement pattern. The specific process of generating tree cell identification data is as follows;

[0073] In the cell feature extraction submodule, the cell structure in the image is analyzed based on the growth ring pattern analysis information, the size and shape of the cells are measured, the geometric parameterization method is used to describe the morphology of each cell, the arrangement pattern between cells is analyzed through the spatial arrangement algorithm, and the organizational structure of the cell group is identified. The cell recognition model is used, and the formula is: ,in, Represents the difference in cell structure, For the The size parameters of the cells, Generate tree cell identification data for the shape factor.

[0074] The texture trend analysis submodule analyzes the annual ring texture characteristics of various trees with different tree ages based on tree cell identification data, identifies texture change trends and characteristics, and generates a tree texture feature set. The specific process is as follows:

[0075] In the texture trend analysis submodule, tree cell identification data is used to compare and analyze the annual ring textures of trees of different ages and species. The texture change parameters of trees of different ages are calculated through statistical analysis methods, and the trend of texture change is identified and quantified. The trend analysis model is adopted, and the formula is: ,in, Represents the texture trend indicator, For the Texture feature values ​​of tree ages, is the texture variation coefficient of the target age group, For the number of samples, a tree texture feature set is generated.

[0076] See also Figure 2 and Figure 4 , the environmental impact factor analysis module includes:

[0077] The environmental data collection submodule collects environmental data related to tree growth based on the tree texture feature set, including the average temperature and humidity of the area, and light intensity. The specific process of generating the tree area environmental analysis results is as follows;

[0078] In the environmental data acquisition submodule, based on the tree texture feature set, the data of the meteorological station and the light sensor in the specified area are acquired in real time, the data collected by each sensor is cleaned and formatted, invalid data and outliers are removed, and the environmental factors related to tree growth are monitored. The environmental impact model is used, and the formula is ,in, is the average temperature, is the average humidity, is the average light intensity, is the weight coefficient of temperature, is the weight coefficient of humidity, is the light weight coefficient, It is an environmental adaptability index used to evaluate the impact of current environmental conditions on the growth of target tree species and generate tree regional environmental analysis results.

[0079] The influencing factor analysis submodule evaluates the impact of temperature changes and light intensity on tree growth based on the results of tree regional environmental analysis. The specific process of generating environmental interaction evaluation data is as follows:

[0080] In the influencing factor analysis submodule, based on the results of the tree regional environmental analysis, the effects of various environmental factors including temperature changes and light intensity on tree growth are evaluated. The statistical regression method is used to calculate the influence of various factors on the growth rate and health status of trees. The environmental impact assessment model is adopted, and the formula is: ,in, Represents the total impact value of environmental factors, For the The influence coefficient of environmental factors is Generate environmental interaction assessment data for the corresponding environmental parameter values.

[0081] The environmental characteristics integration submodule analyzes the impact of various environmental factors on annual ring characteristics based on environmental interaction assessment data, and identifies the relationship with tree density and annual ring width. The specific process of generating environmental response characteristics analysis results is as follows;

[0082] In the environmental characteristics integration submodule, based on the environmental interaction assessment data, the effects of various environmental factors including temperature, humidity and light intensity on the formation and characteristic changes of annual rings were analyzed. Referring to the association between the target factors and tree density and annual ring width, a multivariate analysis model was used, and the formula was: ,in, represents the environmental response characteristic index, For the The evaluation value of an environmental characteristic, is the influence coefficient of the annual ring characteristics associated with the environmental characteristics, Generate environmental response characteristic analysis results for the total number of reference environmental factors.

[0083] See also Figure 2 and Figure 5 , the spectral feature extraction and analysis module includes:

[0084] The spectral data acquisition submodule performs spectral scanning of trees at multiple wavelengths based on the environmental response characteristic analysis results, analyzes the reflection and absorption data of multiple wavelengths, and generates a multi-spectral scanning data set. The specific process is as follows;

[0085] In the spectral data acquisition submodule, based on the environmental response characteristic analysis results, spectral scanning of trees at multiple wavelengths is performed, and the spectral scanning instrument system is used to collect reflection and absorption data, including a wide range of wavelengths from ultraviolet to infrared. Through the spectral data integration model, the formula is ,in, represents the spectral comprehensive index, is the wavelength index, is the total number of wavelengths scanned, is the wavelength The weight coefficient of It is the reflection or absorption data corresponding to the wavelength, generating a multi-spectral scanning data set.

[0086] The spectral response analysis submodule analyzes the intensity of spectral reflection and absorption of multiple bands based on the multi-spectral scanning data set, identifies the characteristic points of spectral data, and generates the specific process of spectral information analysis results as follows;

[0087] In the spectral response analysis submodule, based on the multi-spectral scanning data set, the reflection and absorption intensities of multiple band spectra are analyzed, and the spectral feature analysis model is used to identify the key feature points in the spectral data, including peaks and valleys. The spectral feature recognition model is used, and the formula is: ,in, represents the spectral feature identification index, It is The strength of the feature points, is the coefficient of variation of the target feature points, is the number of characteristic points, generating the spectral information analysis result.

[0088] Based on the spectral information analysis results, the spectral feature encoding submodule converts the analyzed spectral features, including spectral peaks, absorption bands and reflectance, into digital vectors, analyzes the spectral response characteristics of various tree types, and generates the tree spectral classification vector information in the following specific process;

[0089] In the spectral feature encoding submodule, based on the spectral information analysis results, the spectral features are converted into digital vectors. The spectral peak, absorption band and reflectance parameters of each tree type are encoded into digital vectors through the spectral encoding algorithm. The spectral vector encoding model is adopted, and the formula is: ,in, represents the tree spectral classification vector, is the spectral eigenvector eigenvalues, is the encoding coefficient of the target feature, is the total number of features, generating tree spectrum classification vector information.

[0090] See also Figure 2 and Figure 6 , the tree type analysis and identification module includes:

[0091] The tree species spectrum comparison submodule uses the tree spectrum classification vector information to match with the historical tree species spectrum data, evaluates the matching degree between the spectrum vectors of various trees and the historical data, and generates the spectrum data matching results in the following specific process:

[0092] In the tree species spectrum comparison submodule, the tree spectrum classification vector information is used to compare with the historical tree species spectrum database, and the vector similarity algorithm is used to evaluate the matching degree between the spectrum vectors of various trees and historical data. The spectrum similarity evaluation model is adopted, and the formula is: ,in, represents the matching score, is the current tree spectrum vector, is the historical spectrum vector, It is the vector dimension, generating the spectral data matching result.

[0093] The tree species feature identification submodule analyzes the spectral peaks and key wavelength ranges of various trees through spectral data matching results, identifies the spectral fingerprints of various trees and records the tree species characteristics. The specific process of generating tree species characteristic analysis information is as follows;

[0094] In the tree species feature identification submodule, the spectral peaks and key wavelength ranges of various trees are analyzed through spectral data matching results, and the spectral peak recognition algorithm is used to identify the spectral fingerprint of each tree. The spectral fingerprint analysis model is used, and the formula is: ,in, represents the intensity index of the tree species spectral fingerprint, It is The intensity of the spectral peak, is the weight of the spectral peak, is the number of peaks considered to generate information on the tree species characteristics analysis.

[0095] The tree species classification result integration submodule builds a tree species feature database based on the tree species characteristic analysis information, including the tree spectral fingerprint, annual ring texture characteristics and annual ring density change characteristics. The specific process of generating the tree species identification feature information set is as follows;

[0096] In the tree species classification result integration submodule, a tree species feature database is constructed based on the tree species characteristic analysis information, including the tree spectral fingerprint, annual ring texture characteristics and annual ring density change characteristics. Through the data integration model, the formula is: ,in, represents the comprehensive score of tree species characteristics, It is The score of the feature, is the feature importance weight, is the number of features, generating a tree species identification feature information set.

[0097] See also Figure 2 and Figure 7 , the climate characteristics impact analysis module includes:

[0098] The climate data analysis submodule analyzes historical climate data associated with tree growth based on the tree species identification feature information set, including temperature change patterns, humidity change patterns, annual precipitation, and periodic change patterns of climate conditions. The specific process of generating historical meteorological information analysis results is as follows;

[0099] In the climate data analysis submodule, based on the tree species identification feature information set, the historical climate data associated with tree growth are analyzed, including temperature change patterns, humidity change patterns and annual precipitation. The time series analysis method is used to identify the periodic change pattern of climate conditions. The climate cycle analysis model is adopted, and the formula is ,in, represents the climate periodicity index, represents the time series data of temperature, Represents time series data of humidity, represents the time series data of precipitation, represents the fast Fourier transform, is the temperature coefficient, is the coefficient of humidity, It is the coefficient of precipitation, which generates the analysis results of historical meteorological information.

[0100] The tree ring impact assessment submodule is based on the analysis results of historical meteorological information, analyzes the relationship between climate data of multiple years and the change of tree ring width and density, identifies the impact of climate factors on tree ring growth, and generates the specific process of climate factor impact assessment information as follows;

[0101] In the tree ring impact assessment submodule, based on the analysis results of historical meteorological information, the relationship between climate data of multiple years and the change of tree ring width and density is analyzed. Through correlation analysis and regression model, the impact of climate factors on tree ring growth is determined. The climate impact regression model is used, and the formula is: ,in, Indicates the annual ring change index, is the temperature influence coefficient, is the influence coefficient of humidity, is the influence coefficient of precipitation, , , is the regression coefficient, generating information for the impact assessment of climate factors.

[0102] The growth factor assessment submodule analyzes the impact of climate change on the growth patterns of various tree rings based on the climate factor impact assessment information, including ring width, tree density, and ring color. The specific process of generating the analysis results of the tree ring growth pattern is as follows;

[0103] In the growth factor assessment submodule, based on the climate factor impact assessment information, the impact of climate change on the growth patterns of various tree rings, including ring width, tree density, and ring color, was analyzed using a multivariate impact model with the formula: ,in, represents the tree ring growth pattern index, is the influence function, is the result of the climate factor impact assessment. represents the influence factor of the annual ring width, represents the influencing factor of annual ring density, Indicates the influencing factor of the annual ring color, , , is the adjustment factor that generates the results of the tree ring growth pattern analysis.

[0104] See also Figure 2 and Figure 8 ,The tree age prediction model building modules include:

[0105] The tree age data preparation submodule analyzes the historical growth records of trees and the corresponding climate data based on the results of the tree ring growth pattern analysis, including tree ring width, tree density, tree height, precipitation, and temperature changes. The specific process of generating the tree age modeling data set is as follows;

[0106] In the tree age data preparation submodule, based on the analysis results of the annual ring growth pattern, the historical growth records of trees are analyzed. Combined with the corresponding climate data including annual ring width, tree density, tree height, precipitation, and temperature changes, data integration and feature extraction techniques are used to construct a tree age modeling dataset. The model is constructed using the dataset. The formula is: ,in, represents the tree age modeling dataset, It is climate or growth characteristic values, is the weight coefficient of the target feature, is the number of features, generating a tree age modeling dataset.

[0107] The growth factor analysis submodule analyzes the impact of climate conditions on tree ring formation based on the tree age modeling dataset, identifies key climate factors that affect tree ring growth, and generates the specific process of climate influencing factor identification information as follows;

[0108] In the growth factor analysis submodule, based on the tree age modeling data set, the effects of climate conditions including temperature, humidity, and precipitation on tree ring formation are analyzed. Statistical analysis methods are used to determine the factors that have a key impact on tree ring growth. Through the analysis model, the formula is: ,in, represents the index of climate impact factors, It is The measured values ​​of the climate factors, is the coefficient of influence on the annual ring, is the number of climate factors considered, generating identification information of climate influencing factors.

[0109] The age model construction submodule uses the random forest algorithm to build a tree age prediction model based on the identification information of climate influencing factors, and uses tree ring images and environmental factors to predict tree age. The specific process of generating tree age identification information is as follows;

[0110] In the age model construction submodule, based on the identification information of climate influencing factors, the random forest algorithm is used to build a tree age prediction model. By analyzing the tree ring images and environmental factors, including tree ring width, density, and key climatic conditions, the age of the trees is predicted and the tree age identification information is generated.

[0111] Random forest algorithm, according to the formula:

[0112] ;

[0113] Calculate the tree age, where is the index used to traverse all decision tree models, is the total number of decision trees, For the The weight of the decision tree model, is the input tree ring image data, To predict tree age, For the A decision tree takes into account environmental parameters , , The prediction results after is the environmental stability parameter, is the light intensity, is the coefficient of variation of precipitation.

[0114] The specific execution process of the formula is as follows:

[0115] The tree ring image data are preprocessed to extract key texture features. Tree ring image data, environmental stability parameters, light intensity, and precipitation variation coefficient data are collected and input into the random forest model for calculation. By summarizing the outputs of multiple decision trees and using the weighted average method, the weights are determined by the relevance of historical data, and the tree age prediction results are calculated to identify the ages of multiple trees.

[0116] See also Figure 2 and Fig. 9A method for identifying tree age based on image recognition is provided. The method for identifying tree age based on image recognition is performed based on the above-mentioned tree age identification system based on image recognition, and comprises the following steps:

[0117] S1: Based on tree ring image data, analyze the texture characteristics of tree rings, including cell structure differences and texture patterns, and generate tree ring texture analysis information;

[0118] S2: Using tree ring texture analysis information, analyze the impact of various external environments on tree ring shape and growth, including precipitation and light intensity, and generate environmental regulation texture impact data;

[0119] S3: Based on the environmental adjustment texture impact data, the spectral reflectance and absorptivity of various trees are collected to perform spectral response analysis of the trees, identify the spectral characteristics of the tree species, and generate a spectral response feature set;

[0120] S4: Use the spectral response feature set to match and classify tree species by comparing with the spectral database of known tree species, and generate tree species feature classification results;

[0121] S5: Based on the classification results of tree species characteristics, evaluate the long-term impact of climate change on the growth patterns of tree rings of various trees, build a dynamic tree age prediction model, and combine the basic information of tree rings, including the number of tree rings and effective sample length, to generate tree age identification information.

[0122] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A tree age identification system based on image recognition, characterized in that: The system comprises: The image texture analysis and recognition module analyzes the texture features in tree ring images based on tree ring image data, including differences in cell structures, and generates a tree texture feature set; The environmental impact factor analysis module analyzes the influence of humidity, temperature, and light on the shape of annual rings based on the tree texture feature set, and generates an environmental response feature analysis result; The spectral feature extraction and analysis module collects the spectral response information of trees based on the environmental response feature analysis result, analyzes the response characteristics of various trees, and generates tree spectral classification vector information; The tree type analysis and identification module uses the tree spectral classification vector information to identify tree species characteristics and classify tree species by analyzing the characteristic intensities and distribution patterns of multiple spectral vectors, thereby generating a tree species identification feature information set; The climate characteristic impact analysis module evaluates the growth pattern of tree rings under differentiated climate conditions based on the tree species identification characteristic information set, analyzes the impact of climate change on tree ring width and density, and generates tree ring growth pattern analysis results; The tree age prediction model building module builds a tree age dynamic prediction model based on the analysis results of the annual ring growth pattern, combined with geographical and climatic factors, and generates tree age identification information by combining the basic information of the annual rings identified by the image texture analysis and recognition module, including the number of annual rings and effective sample length; The tree age prediction model building module includes: The tree age data preparation submodule analyzes the historical growth records of trees and corresponding climate data, including growth ring width, tree density, tree height, precipitation, and temperature changes, based on the analysis results of the tree ring growth pattern, and generates a tree age modeling data set; The growth factor analysis submodule analyzes the influence of climate conditions on tree ring formation based on the tree age modeling data set, identifies key climate factors that affect tree ring growth, and generates climate influencing factor identification information; The age model building submodule uses a random forest algorithm based on the climate influencing factor identification information to build a tree age prediction model, uses tree ring images and environmental factors to predict tree age, and generates tree age identification information; The random forest algorithm is based on the formula: ； Calculate the tree age, where is the index used to traverse all decision tree models, is the total number of decision trees, For the The weight of the decision tree model, is the input tree ring image data, To predict tree age, For the A decision tree takes into account environmental parameters , , The prediction results after is the environmental stability parameter, is the light intensity, is the coefficient of variation of precipitation.

2. The tree age identification system based on image recognition according to claim 1 is characterized in that: The tree texture feature set includes cell wall thickness information, cell arrangement pattern information, and cell size variation information. The environmental response feature analysis results include analysis results of the impact of light cycle on annual ring shape, analysis results of the impact of humidity changes on annual ring density, and analysis results of the impact of temperature fluctuations on annual ring color. The tree spectral classification vector information specifically includes tree species spectral absorption peak information, tree species spectral reflection peak information, and tree species spectral intensity distribution information. The tree species identification feature information set includes tree species spectral fingerprint data, tree species morphological characteristics, and tree species ecological preferences. The annual ring growth pattern analysis results include the impact of temperature increase on annual ring thickness, the impact of annual average precipitation on annual ring density, and the impact of seasonal temperature difference on annual ring growth rate. The tree age identification information specifically includes tree age range prediction information, growth start year prediction information, and tree age prediction error range information.

3. The tree age identification system based on image recognition according to claim 1 is characterized in that: The image texture analysis and recognition module comprises: The texture feature recording submodule analyzes the texture details of multiple tree ring images based on tree ring image data, identifies the continuity, fracture pattern and color information of the texture, and generates tree ring texture analysis information; The cell feature extraction submodule analyzes the cell structure differences in the tree ring image based on the tree ring pattern analysis information, including cell size, shape and mutual arrangement pattern, and generates tree cell identification data; The texture trend analysis submodule analyzes the annual ring texture features of various trees with different tree ages based on the tree cell identification data, identifies texture change trends and features, and generates a tree texture feature set.

4. The tree age identification system based on image recognition according to claim 1, characterized in that: The environmental impact factor analysis module includes: The environmental data collection submodule collects environmental data associated with tree growth based on the tree texture feature set, including average temperature and humidity and light intensity of the area, and generates tree area environmental analysis results; The influencing factor analysis submodule evaluates the impact of temperature changes and light intensity on tree growth based on the tree area environmental analysis results, and generates environmental interaction evaluation data; The environmental characteristic integration submodule analyzes the impact of various environmental factors on annual ring characteristics based on the environmental interaction assessment data, identifies the relationship with tree density and annual ring width, and generates environmental response characteristic analysis results.

5. The tree age identification system based on image recognition according to claim 1 is characterized in that: The spectral feature extraction and analysis module comprises: The spectral data acquisition submodule performs spectral scanning of trees at multiple wavelengths based on the environmental response characteristic analysis results, analyzes the reflection and absorption data of the multiple wavelengths, and generates a multi-spectral scanning data set; The spectral response analysis submodule analyzes the intensity of spectral reflection and absorption of multiple bands based on the multi-spectral scanning data set, identifies the characteristic points of spectral data, and generates spectral information analysis results; The spectral feature encoding submodule converts the analyzed spectral features, including spectral peaks, absorption bands and reflectance, into digital vectors based on the spectral information analysis results, analyzes the spectral response characteristics of various tree types, and generates tree spectral classification vector information.

6. The tree age identification system based on image recognition according to claim 1, characterized in that: The tree type analysis and identification module includes: The tree species spectrum comparison submodule uses the tree spectrum classification vector information to match with the historical tree species spectrum data, evaluates the matching degree between the spectrum vectors of multiple trees and the historical data, and generates spectrum data matching results; The tree species characteristic identification submodule analyzes the spectral peaks and key wavelength ranges of various trees through the spectral data matching results, identifies the spectral fingerprints of various trees and records the tree species characteristics, and generates tree species characteristic analysis information; The tree species classification result integration submodule constructs a tree species feature database based on the tree species characteristic analysis information, including the tree's spectral fingerprint, annual ring texture characteristics and annual ring density change characteristics, and generates a tree species identification feature information set.

7. The tree age identification system based on image recognition according to claim 1 is characterized in that: The climate characteristics impact analysis module includes: The climate data analysis submodule analyzes historical climate data associated with tree growth based on the tree species identification feature information set, including temperature change patterns, humidity change patterns, and annual precipitation, analyzes the periodic change patterns of climate conditions, and generates historical meteorological information analysis results; The tree ring impact assessment submodule analyzes the relationship between the climate data of multiple years and the change of tree ring width and density based on the historical meteorological information analysis results, identifies the impact of climate factors on tree ring growth, and generates climate factor impact assessment information; The growth factor assessment submodule analyzes the impact of climate change on the growth patterns of various tree rings, including tree ring width, tree density, and tree ring color, based on the climate factor impact assessment information, and generates a growth ring growth pattern analysis result.

8. A tree age identification method based on image recognition, characterized in that: The tree age identification system based on image recognition according to any one of claims 1 to 7 is implemented, comprising the following steps: Based on tree ring image data, the texture characteristics of tree rings are analyzed, including differences in cell structure and texture patterns, to generate tree ring texture analysis information; Utilizing the tree ring texture analysis information, analyzing the effects of various external environments on tree ring shape and growth, including precipitation and light intensity, to generate environmental conditioning texture impact data; Based on the environmental adjustment texture impact data, the spectral reflectance and absorptivity of various trees are collected, the spectral response analysis of the trees is performed, the spectral characteristics of the tree species are identified, and a spectral response characteristic set is generated; Using the spectral response feature set, by comparing with a spectral database of known tree species, matching and classifying tree species, and generating a tree species feature classification result; Based on the classification results of tree species characteristics, the long-term impact of climate change on the growth patterns of tree rings of various trees was evaluated, and a dynamic tree age prediction model was constructed. Combined with basic tree ring information, including the number of tree rings and effective sample length, tree age identification information was generated.

Images