A historical forestry field survey sample migration method based on forest remote sensing classification

Abstract

The present invention provides a method for migrating samples of historical forestry field surveys based on forest remote sensing classification, which belongs to the field of forest monitoring. The method first determines the monitoring area, historical years and target years, obtains historical sample points of forestry field surveys in historical years and target sample points of forestry field surveys in target years within the monitoring area, then obtains long-term multi-source remote sensing data of the historical sample points from the historical years to the target years and characterizes the forest growth succession characteristics of the historical sample points, then determines whether the forest type to which the historical sample points belong has changed from the historical years to the target years, and migrates the forestry field survey sample data of the historical sample points that have not changed to the target year. The present invention utilizes the growth succession characteristics from the historical years to the target years to migrate the data of historical forestry field survey sample points where no forest type change has occurred to the target year, thereby improving the migration efficiency and quality of the samples and increasing the number of samples in the target year.

Description

Technical Field

[0001] The present invention belongs to the field of forest monitoring, and in particular relates to a historical forestry field survey sample migration method based on forest remote sensing classification. Background Art

[0002] Mountain forests are rich in variety and serve as the backbone of ecological security barriers, important reservoirs of natural resources, and a treasure trove of biodiversity. Accurate classification and mapping of mountain forest types is of great scientific significance for sustainable forest resource management, biodiversity conservation, and ecosystem service evaluation. Traditional mountain forest classification relies on field survey data. However, with the development of satellite remote sensing technology, it has become the primary means of mapping mountain forest types due to its wide coverage, short revisit cycles, and low data acquisition costs.

[0003] In recent years, remote sensing has been used to classify mountain forests, often using traditional machine learning and deep learning methods. However, the accuracy of these methods is heavily dependent on the quality and quantity of real-world samples. Mountainous areas are characterized by complex terrain, variable environmental factors, and limited accessibility, making it difficult to obtain large numbers of field forestry survey samples. Consequently, these methods often suffer from low classification accuracy and significant spatial variations in confidence when used in remote sensing classification of mountain forest types.

[0004] Existing techniques typically increase the sample size and improve remote sensing classification accuracy by migrating historical forestry field survey samples to the target year. Historical national forestry field survey data contains a large number of known historical samples. Migrating this information to the target year to increase the sample size would fundamentally address the small sample size issue in remote sensing classification of mountain forest types. However, existing sample migration methods are immature, inefficient, and of poor quality, resulting in limited improvement in remote sensing classification accuracy. Summary of the Invention

[0005] In view of the above-mentioned defects or deficiencies in the prior art, the present invention aims to provide a method for migrating historical forestry field survey samples based on forest remote sensing classification, which utilizes the growth succession characteristics from the historical year to the target year to migrate the historical forestry field survey sample point data where no forest type replacement has occurred to the target year, so as to expand the number of samples in the target year and improve the migration efficiency and quality of the samples.

[0006] In order to achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0007] An embodiment of the present invention provides a method for migrating historical forestry field survey samples based on forest remote sensing classification, the method comprising the following steps:

[0008] Step S1, determining the monitoring area, historical years and target years, obtaining historical sample points of forestry field surveys in historical years and target sample points of forestry field surveys in the target year within the monitoring area;

[0009] Step S2, obtaining long-term multi-source remote sensing data of the historical sample points from the historical year to the target year;

[0010] Step S3, using long-term multi-source remote sensing data to characterize the forest growth succession characteristics of the historical sample points; the forest growth succession characteristics include spectral similarity of the historical sample points in the historical year and the target year, the temporal trajectory of the historical sample points from the historical year to the target year, and the environmental similarity between the historical sample points and the target sample points of the same type;

[0011] Step S4: Based on the forest growth succession characteristics, determine whether the forest type of the historical sample point has changed from the historical year to the target year; if the three characteristics of the historical sample point are similar in terms of spectral similarity between the historical year and the target year, time series trajectory between the historical year and the target year, and geographical environment similarity between the historical sample point and the sample point of the same type in the target year, then the forest type to which the historical sample point belongs has not changed; otherwise, it is determined that a change has occurred;

[0012] Step S5: Migrate forestry field survey sample data of historical sample points that have not been replaced to the target year, and delete historical sample points that have been replaced.

[0013] As a preferred embodiment of the present invention, step S1 obtains forestry field survey sample point data of historical years and target years in the monitoring area. First, the forestry field survey data of the monitoring area in historical years and target years are obtained, and then the surface-to-point method is used to generate historical sample point data and target sample point data maps respectively.

[0014] As a preferred embodiment of the present invention, the long-term multi-source remote sensing data in step S2 includes all available Landsat Collection2Tier1 surface reflectance data images, Sentinel-1 / 2 data, SRTM-DEM data, climate data and soil moisture data on the Google Earth Engine GEE platform corresponding to the period from historical years to target years.

[0015] As a preferred embodiment of the present invention, step S4 determines whether the forest type of the historical sample point has changed from the historical year to the target year by using a threshold segmentation method.

[0016] As a preferred embodiment of the present invention, the formula of the threshold segmentation method is as follows:

[0017]

[0018] And, Environmental similarity=max(S i1 ,S i2 ,…,S ij )

[0019] In formula (8), Foresttype is the forest type label of the historical sample point; SAD and ED are spectral similarity indicators that quantify the historical sample point between the historical year and the target year, σ1 and σ2 are the thresholds of SAD and ED respectively; disturbanceyear is the forest disturbance detection result of the historical sample point from the historical year to the target year, and Historical year is the historical year; Environmental similarity is the environmental similarity between the historical sample point and the target sample point of the same type, σ3 is the segmentation threshold of environmental similarity, and S i,j is the environmental similarity between all sample points of the same forest type in the historical sample point i and the target sample point j.

[0020] As a preferred embodiment of the present invention, the value of σ1 is 0.95; the value of σ2 is 0.1.

[0021] As a preferred embodiment of the present invention, the value of σ3 is determined according to actual conditions.

[0022] As a preferred embodiment of the present invention, the migration model of the forestry field survey sample data of the historical sample points in step S5 is:

[0023]

[0024] In formula (9), forest type 历史样本点 For each historical forestry field survey sample, the forest type of the historical sample point from the historical year to the target year is taken. If the forest type has not been replaced at the historical sample point, it means that the forest type label of the sample point has not changed and it is migrated to the target year for forest type remote sensing classification. Otherwise, the type label information of the sample point has been replaced and the sample point is deleted.

[0025] The technical solution provided by the embodiment of the present invention has the following beneficial effects:

[0026] The present invention provides a method for migrating historical forestry field survey samples based on forest remote sensing classification. Based on the growth succession information of historical samples between the historical year and the target year, historical samples that have not undergone forest type change are migrated to the target year to achieve remote sensing classification of forest types in the target year. It is well known that even in the absence of natural or human interference, forest communities continue to regenerate, grow, and succeed over time. This growth succession process is often accompanied by changes in characteristics such as tree species composition, stand structure, canopy cover, and vegetation functional traits. These changes often lead to changes in the spectral characteristics of the forest canopy and the local microclimate. For example, in the pioneer stage (such as after a forest fire or glacial retreat), stand patches are primarily dominated by fast-growing tree species (such as poplar and birch). In the secondary and mature forest stages, these are gradually replaced by taller, shade-tolerant tree species (such as oak and maple), gradually forming a stable stand structure and forest ecosystem. Overmature forests may contain endangered or environmentally adapted tree species, resulting in a more complex and diverse stand structure. The present invention overcomes the problem that existing methods ignore the impact of forest growth succession on sample migration quality, can effectively improve the migration quality of historical forestry field survey samples, and improve the classification accuracy of forest types in the target year.

[0027] Of course, it is not necessary to achieve all of the advantages described above simultaneously in order to implement any product or method of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 This is a flow chart of a method for migrating historical forestry field survey samples based on forest remote sensing classification provided by an embodiment of the present invention;

[0030] Figure 2 is a spatial distribution pattern map of forest types at altitude in the monitoring area in 2023, inferred by applying the sample migration method in an embodiment of the present invention;

[0031] Figure 3 This is a comparison chart of PA indicators for forest remote sensing classification using the sample migration method in the embodiment of the present invention and other existing methods;

[0032] Figure 4 This is a comparison chart of UA indicators for forest remote sensing classification using the sample migration method in the embodiment of the present invention and other existing methods;

[0033] Figure 5 This is a comparison chart of the F1 scores of forest remote sensing classification using the sample migration method in the embodiment of the present invention and other existing methods;

[0034] Figure 6 This is a comparison chart of the number of migrated samples of different forest types when applying the sample migration method in the embodiment of the present invention and other existing sample migration methods. DETAILED DESCRIPTION

[0035] After discovering the above-mentioned issues, the inventors of this application conducted a thorough study of existing methods for migrating historical forestry field survey samples. This study found that several methods exist for migrating historical field survey samples for target year land cover classification, including historical sample reuse, historical sample weighting, and dual-temporal imagery change detection. However, all methods have certain shortcomings.

[0036] For example, the historical sample weighting method dynamically adjusts sample weights based on the contribution of historical samples to the target model and uses the weighted historical samples for land cover classification in the target year. For example, Zhou et al. (2016) introduced a transfer term into the target classification model. Based on the classification accuracy of historical and target samples in the model, they iteratively adjusted sample weights (if a historical sample is misclassified, it indicates that the sample's spatial distribution is inconsistent with the target year's remote sensing data, and its weight should be reduced; if a target sample is misclassified, it indicates that the sample has not been fully learned by the model, and its weight should be increased) until all target samples are correctly classified. The weighted samples are then used for land cover classification. Compared with the historical sample reuse method, this method reduces the impact of historical samples that are spatially inconsistent with the target year's remote sensing data on classification accuracy, but it does not account for the potential impact of changes in historical sample label information caused by land cover changes on the target year's land cover classification accuracy.

[0037] Another example is the dual-temporal image change detection method, which assumes that changes in surface cover types will cause changes in surface radiation characteristics; first, the spectral feature similarity between the historical and target year remote sensing data is calculated, and then a similarity threshold is set to determine whether the label information of the historical samples has changed, and the historical samples whose label information has not changed are migrated to the target year; finally, based on the migrated samples and the remote sensing data of the target year, a machine learning algorithm is used to realize the surface cover classification of the target year. For example, Lin et al. (2019) migrated historical samples with no label information changes to the target year by comparing the spectral feature differences (Change Vector Analysis, CVA) between historical samples (2010) and the target year, and used the migrated samples and Landsat data to realize the automatic classification of surface cover types in rapidly urbanized areas; for example, Huang et al. (2020) based on the global surface cover type sample dataset (Firstall-season sample set) collected by the FROM-GLC project in 2015, migrated historical samples with no label information changes to the target year by calculating the spectral similarity and spectral distance between the target year and the 2015 Landsat data, and used these migrated samples and Landsat data to perform global surface cover classification in 1990, 1995, 2000, 2005 and 2010 respectively. Although this type of method reduces the impact of the replacement of historical sample label information on the quality of migrated samples to a certain extent, it still has the following two limitations in the process of migrating historical forestry field survey samples: 1) The spectral characteristics of different forest types in multispectral remote sensing images are relatively small. Using only dual-phase remote sensing images can usually only identify the replacement of forests to non-forests, and it is difficult to effectively identify the replacement information of forest types at the stand scale, such as historical samples from coniferous / broad-leaved forests to mixed coniferous and broad-leaved forests; 2) This method does not fully consider the potential impact of the natural growth succession of forests from the historical year to the target year on the quality of sample migration.

[0038] The above analysis shows that the historical sample weighting method is cumbersome and requires a tedious dynamic weight adjustment process to generate samples suitable for the target year's land cover classification. Although the dual-temporal remote sensing image change detection method has been improved based on the reuse of historical samples, these methods do not consider the impact of the natural growth succession of the forest between the historical year and the target year on the quality of sample migration, resulting in the low quality of migrated forestry field survey samples. It is well known that even in the absence of natural or human interference, forest communities will continue to renew, grow, and succeed over time; this growth and succession process is usually accompanied by changes in characteristics such as tree species composition, stand structure, canopy cover, and vegetation functional traits, and these characteristic changes often cause changes in the forest canopy spectral characteristics and local microclimate. For example, in the pioneer stage (such as after a forest fire or after glacier retreat), forest stand patches are mainly dominated by fast-growing tree species (such as poplar and birch); in the secondary forest and mature forest stages, they are gradually replaced by taller, shade-tolerant tree species (such as oak and maple), gradually forming a stable forest stand structure and forest ecosystem; the overmature forest stage may contain endangered or environmentally adapted tree species, and the forest stand structure becomes more complex and diverse. Therefore, in the process of migrating historical forestry field survey samples, it is necessary to comprehensively consider the impact of forest growth succession on the quality of sample migration, so as to improve the quality of migrated samples, increase the number of effective samples in the target year, and improve the accuracy of remote sensing classification of mountain forest types under small sample conditions. However, the monitoring methods of existing technologies do not take the above issues into account.

[0039] It should be noted that the defects existing in the solutions in the above-mentioned prior art are the results obtained by the inventors after practice and careful research. Therefore, the discovery process of the above-mentioned problems and the solutions proposed in the embodiments of the present invention below for the above-mentioned problems should all be the contributions made by the inventors to the present invention in the process of the invention.

[0040] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Generally, the components of the embodiments of the present invention described and shown in the accompanying drawings can be arranged and designed in various different configurations. It should be noted that the embodiments of the present invention and the features in the embodiments can also be combined with each other without conflict.

[0041] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0042] After the above in-depth analysis, an embodiment of the present invention provides a method for migrating historical forestry field survey samples based on forest remote sensing classification. This method utilizes the growth succession information of historical samples from the historical year to the target year to migrate historical samples that did not undergo forest type replacement in the same monitoring area to the target year, thereby expanding the number and quality of samples for forest type remote sensing classification. First, multi-source remote sensing data from the historical year to the target year is used to quantitatively characterize the growth succession characteristics of the forest. This growth succession characteristic is then used to determine whether the forest type label of the historical sample has undergone a replacement. Finally, based on the judgment result of whether a replacement has occurred, historical samples that have not undergone a replacement are migrated to the target year, thereby expanding the number and quality of samples for forest remote sensing classification and improving the accuracy and precision of remote sensing classification.

[0043] like Figure 1 As shown, the historical forestry field survey sample migration method based on forest remote sensing classification includes the following steps:

[0044] Step S1, determining the monitoring area, historical years and target years, obtaining historical sample points of forestry field surveys in historical years and target sample points of forestry field surveys in target years within the monitoring area.

[0045] In this step, the forestry field survey sample point data of the historical years and the target years in the monitoring area are obtained. First, the forestry field survey data of the monitoring area in the historical years and the target years are obtained, and then the historical sample point data and the target sample point data are generated respectively using the surface-to-point method.

[0046] Step S2, obtaining long-term multi-source remote sensing data of the historical sample points from the historical year to the target year.

[0047] In this step, the long-term multi-source remote sensing data include all available Landsat Collection 2Tier1 surface reflectance data images, Sentinel-1 / 2 data, SRTM-DEM data, climate data and soil moisture data on the Google Earth Engine GEE platform corresponding to the period from historical years to target years.

[0048] Step S3, using long-term multi-source remote sensing data to characterize the forest growth succession characteristics of historical sample points; the forest growth succession characteristics include the spectral similarity of the historical sample points in the historical year and the target year, the time series trajectory of the historical sample points from the historical year to the target year, and the environmental similarity between the historical sample points and the target sample points of the same type.

[0049] In this step, the spectral similarity between historical sample points in historical years and target years can be calculated using methods such as spectral angular distance (SAD) and Euclidean distance (ED); the time series trajectory of historical sample points from historical years to target years can be calculated using the improved LandTrendr algorithm (iLandTrendr); the environmental similarity between historical sample points and target sample points of the same type can be calculated using methods such as random forest regression algorithm, Gaussian similarity function, and the relationship between various geographical environmental covariates.

[0050] In a specific embodiment, the spectral similarity between historical sample points in the historical year and the target year is calculated, Landsat data of the monitoring area in the historical year and the target year are collected, and the spectral similarity between the optical remote sensing images of the historical samples in the historical year and the target year is quantified using two indicators: spectral angle distance (SAD) and Euclidean distance (ED). Specifically, the steps include:

[0051] Step S311, collecting Landsat data of historical sample points corresponding to historical years and target years;

[0052] Step S312, using SAD and ED to calculate the spectral similarity between the Landsat images of the historical sample points in the historical year and the target year, the calculation formula is as follows:

[0053]

[0054] Where θ is the spectral angle, is the reference spectral feature corresponding to the sample point t1 in historical year, is the spectral signature of the historical sample in the target year t2; i is the corresponding spectral band, which represents bands B1-B5 and B7 in Landsat5TM data or bands B2-B7 in Landsat8 OLI. If the spectral signature remains unchanged between the historical and target years, SAD equals 1 and ED equals 0. Preferably, the segmentation thresholds for SAD and ED are set to 0.95 and 0.1, respectively.

[0055] In another specific embodiment, the method calculates the time series trajectory of historical sample points from the historical year to the target year, collects Landsat data for each vegetation growing season from the historical year to the target year, masks the areas covered by clouds and cloud shadows, and then synthesizes the masked image to generate an image that can well capture the Landsat observations. Finally, based on the long-term Landsat data, the iLandTrendr (Improved Landsat-based detection of trends in disturbance and recovery) algorithm is used to detect whether forest replacement has occurred in the historical sample during this period, that is, to detect whether forest disturbance events such as forest fires, deforestation, pests and diseases have occurred in the historical sample data. Specifically, the following steps are included:

[0056] Step S321 , collecting Landsat data for each vegetation growing season within a long time series, masking areas covered by clouds and cloud shadows, and then synthesizing the masked images to generate an image that can well capture the Landsat observations;

[0057] Step S322: Calculate the normalized burn index (NBR) for each pixel. Then, use linear interpolation to fill in the NBR values ​​corresponding to the years covered by clouds and cloud shadows. Finally, use the SG filtering algorithm to filter out data noise to obtain the NBR time series trajectory of each pixel over time.

[0058] Step S323: Based on the NBR time series trajectory within the long-term year range, the iLandTrendr algorithm is used to detect the year in which forest disturbance occurred in each pixel. That is, assuming that when forest disturbance occurs, the NBR time series trajectory will show a decreasing value, and when the forest recovers, the NBR index will show a gradually increasing trend. Based on the iLandTrendr algorithm, the year corresponding to the node where the NBR value changes from "decline to increase" is considered to be the year in which forest disturbance occurred, and the corresponding pixel is regarded as the pixel in which forest disturbance occurred within the long-term year range.

[0059] In another specific embodiment, the calculation of the environmental similarity between the historical sample points and the target sample points of the same type includes, first, using multi-source remote sensing data of the target year to characterize environmental conditions related to the spatial distribution of forest types, such as forest canopy characteristics (spectral, polarization, phenology, spatial texture, etc.) and geographical environment (precipitation, temperature, soil, etc.); and secondly, using a Gaussian similarity function to compare the similarity of environmental conditions between the historical sample points and the target sample points, specifically including the following steps:

[0060] Step S331, using multi-source remote sensing data acquired in the target year to comprehensively characterize the environmental conditions of each geographical location;

[0061] Step S332: Calculate the environmental similarity between the historical sample point and the target sample point. The calculation formula is as follows:

[0062]

[0063] In formula (4), and are the values ​​of the vth environmental covariate for the historical sample point i and the target sample point j, respectively; m is the number of geographical environmental covariates used to characterize the spatial distribution of forest types. It is based on the forest type samples of the target year. The importance of all classification features in step S2 is evaluated using the random forest algorithm, and features with feature importance greater than 0.00001 are selected to characterize the geographical environment configuration corresponding to each sample point. v (·) is a function used to calculate the geographical similarity between historical sample point i and target sample point j on a single environmental variable v. This paper uses the Gaussian similarity function for calculation, and its calculation formula is shown in Formula 7:

[0064]

[0065] In formula (5), is the standard deviation of the vth environmental covariate in the entire region, It is the square root of the average deviation of the j-th target sample point from all historical sample points of the same type (i = 1, 2, ..., k) on the v-th environmental covariate:

[0066]

[0067] It is a function used to synthesize the similarities of m environmental covariates. The similarities of each environmental covariate are synthesized by using the characteristic importance of each environmental covariate to the forest type classification to obtain the comprehensive environmental similarity between the historical sample point i and the target sample point j:

[0068]

[0069] In formula (7), The Mean Decrease Accuracy (MDA) method in the random forest algorithm is used to evaluate the importance of all features and the feature importance of each environmental covariate for regional forest type classification.

[0070] Step S4: Based on the forest growth succession characteristics, determine whether the forest type of the historical sample point has changed from the historical year to the target year; if the three characteristics of the historical sample point are similar, namely, the spectral similarity between the historical year and the target year, the time series trajectory between the historical year and the target year, and the geographical environment similarity between the historical sample point and the sample point of the same type in the target year, then the forest type to which the historical sample point belongs has not changed; otherwise, it is determined that a change has occurred.

[0071] In this step, the threshold segmentation method is used to determine whether the forest type of the historical sample point has changed from the historical year to the target year. Specifically, the threshold segmentation formula is as follows:

[0072]

[0073] And, Environmental similarity=max(S i1 ,S i2 ,…,S ij )

[0074] In formula (8), Foresttype is the forest type label of the historical sample point; SAD and ED are spectral similarity indicators that quantify the historical sample point between the historical year and the target year, σ1 and σ2 are the thresholds of SAD and ED respectively; the smaller the ED value or the larger the SAD value, the more similar the spectrum is, and the smaller the possibility of forest type replacement; disturbance year is the forest disturbance detection result of the historical sample point from the historical year to the target year. If the year of forest disturbance is less than or equal to the historical year, it means that there is no forest disturbance at the sample location, and the possibility of forest type label information replacement is small; environmental similarity is the environmental similarity between the historical sample point and the target sample point of the same type, σ3 is the segmentation threshold of environmental similarity, S i,j is the environmental similarity between historical sample point i and all sample points of the same forest type in target sample point j. Based on the third law of geography, this paper assumes that similar environmental configurations lead to similar forest types. That is, the greater the environmental similarity between historical sample points and target sample points, the less likely they are to undergo forest type replacement. Therefore, this paper considers each historical sample point collected in the forestry field during the target year as a case study containing a specific "forest type-environmental condition" relationship, representing the appropriate geographical environmental conditions and forest canopy characteristics for that forest type.

[0075] Step S5: Migrate forestry field survey sample data of historical sample points that have not been replaced to the target year, and delete historical sample points that have been replaced.

[0076] In this step, the migration model of forestry field survey sample data of historical sample points is:

[0077]

[0078] In formula (9), forest type 历史样本点 For each historical forestry field survey sample, the forest type of the historical sample point from the historical year to the target year is taken. If the forest type has not been replaced at the historical sample point, it means that the forest type label of the sample point has not changed and can be migrated to the target year for forest type remote sensing classification. Otherwise, the type label information of the sample point has been replaced and the sample point is deleted.

[0079] Yunnan Province in southwestern China is used as the study area. The historical forestry field survey sample migration method based on forest remote sensing classification provided by the embodiment of the present invention is adopted. With 2012 and 2016 as the historical years and 2023 as the target year, the historical samples in the selected area are migrated, and the forest types are remotely sensed classified based on the overall sample data after migration, and the classification results are evaluated.

[0080] First, the historical forestry field survey samples from 2012 and 2016 were migrated; then, based on the migrated samples, remote sensing classification of forest types in 2023 was carried out.

[0081] The monitoring area is located in southwestern China, east of the Himalayas, bordering Myanmar, Vietnam, and Laos. It is a key component of the Indo-Myanmar Biodiversity Hotspot. Forest cover reaches 65% of the area, with relatively young stands, robust forest growth, and significant carbon sequestration potential. The study area ranges from 76.4 to 6,740 meters above sea level, descending in a stepped pattern from north to south. The landforms are complex and diverse, encompassing plains, mountains, hills, and basins. The study area has a subtropical plateau monsoon climate, but due to its complex topography, regional temperature and precipitation show significant spatial heterogeneity, with significant spatial and vertical variations in local climate. The area encompasses a wide range of ecosystems, with the exception of marine and desert regions. The unique and diverse natural environment of the monitoring area provides a suitable habitat for the origin, evolution, and reproduction of diverse organisms. The area is rich in forest resources, encompassing a diverse flora, with Sino-Japanese flora predominating in the east, Sino-Himalayan flora in the west, and paleotropical flora predominating in the south. Accurate forest type classification provides essential data for the sustainable management of regional forest resources.

[0082] In this embodiment, forest type sample point data within the monitoring area are collected, including Yunnan Province forestry field survey data for historical years and target years.

[0083] Among them, the historical sample data includes forestry field survey data from 2012 and 2016 (3120), of which there are 3699 samples in 2012 and 14065 samples in 2016; the forestry field survey samples in the target year are 3210. In this embodiment, the acquisition of forest type samples for each plot is completed in four steps: 1) a series of plots are laid out in Yunnan Province based on comprehensive consideration of spatial representativeness, economy and transportation accessibility; 2) a 30×30 m forest plot is laid out using a leather rope according to the preset longitude and latitude of the plot, and real-time dynamic differential positioning technology (Real Time Kinematic GPS, RTK-GPS) is used to dynamically adjust the actual longitude and latitude information of the plot center and four corner points to determine the plot boundary; 3) for each plot, the diameter at breast height of each tree is measured using a breast diameter ruler, and the forest type to which each tree species with a breast diameter ≥5 cm belongs is manually identified based on the shape of the leaves, and its crown cover area is quantitatively estimated; 4) the forest type to which the plot belongs is determined based on the proportion of crown cover area of ​​each forest type. If the proportion of coniferous forest and broad-leaved forest is less than 65%, the plot belongs to a mixed coniferous and broad-leaved forest; otherwise, it is a pure forest.

[0084] We then acquired long-term, multi-source remote sensing data for the historical sample points from the historical years to the target year. This data included all available Landsat Collection 2 Tier 1 surface reflectance imagery, Sentinel-1 / 2 data, SRTM-DEM data, climate data, and soil moisture data from the Google Earth Engine (GEE) platform, covering the period from the historical forest survey years (2012 / 2016) to the target year (2023).

[0085] After corresponding preprocessing of the above-mentioned long-term multi-source remote sensing data, the forest growth succession characteristics of the historical sample points are characterized.

[0086] The description of forest growth succession characteristics in this step is mainly completed in three steps, including:

[0087] The spectral angular distance (SAD) and Euclidean distance (ED) are used to quantify the spectral similarity between historical samples and the target year. The spectral similarity includes the similarity between the B1-B7 bands of the Landsat data in the historical years (2012 / 2016) and the target year (2023).

[0088] The time series trajectory was quantified using Landsat data from historical samples between the historical year and the target year, and the iLandTrendr algorithm was used to detect whether forest disturbance occurred during this period.

[0089] The environmental similarity between historical sample points in the region and the same type of sample points in each target year was calculated using the Gaussian similarity function and the characteristic importance of each geographic environmental covariate.

[0090] Based on the characteristics of forest growth succession, the forestry field survey sample data corresponding to historical sample points judged to have not undergone forest type replacement will be migrated to the forestry field survey sample data of 2023 to complete the migration of historical data.

[0091] Then, based on the migrated data and the multi-source remote sensing data from 2023, remote sensing classification of forest types for the target year is performed. When performing remote sensing classification of forests, corresponding classification methods can be used, which generally include the following steps:

[0092] Using multi-source remote sensing data from the target year, we extracted classification features related to the remote sensing classification of regional forest types. The extracted classification features included spectrum, polarization, texture, phenology, and environment.

[0093] According to the spatial distribution characteristics of the flora in the monitoring area, the monitoring area is divided into multiple floras, and the optimal classification features for forest type classification of each flora are selected from the relevant classification features using the correlation hierarchical clustering method to reduce the impact of collinearity and redundancy between classification features on the classification results;

[0094] Based on the optimal classification features and forestry field survey sample data of historical sample points that have not undergone replacement and have been moved to the target year, the random forest algorithm is used to classify the forest types of each flora;

[0095] The classification results of forest type classification of all floras were integrated to obtain the spatial distribution map of forest types in the entire monitoring area;

[0096] The mapping accuracy was evaluated using forestry field survey sample data from the target year.

[0097] Other forest remote sensing classification methods in the prior art may also be used.

[0098] When using the above-mentioned flora-based method to perform remote sensing classification of forest types in the selected area, Yunnan Province is divided into multiple floras according to the spatial distribution characteristics of the flora of Yunnan Province. The association hierarchical clustering method is used to select the preferred features that are conducive to the classification of forest types in each plant region to reduce the impact of collinearity and redundancy between classification features on the classification results. The classification features mainly include spectral, polarization, phenological, texture and geographical environment characteristics. The preferred classification features of each flora are shown in Table 1:

[0099] Table 1 Preferred classification characteristics of each flora

[0100]

[0101]

[0102] Figure 2 The spatial distribution map of forest types in the entire monitoring area is obtained by integrating the classification results of 8 floras.

[0103] To verify the effectiveness of the proposed automatic migration method for historical forestry field survey samples based on remote sensing forest classification (a samples transferring method considering forest succession, STM-CFS), this example quantitatively evaluated the accuracy of the proposed method in generating time-series forest canopy height distribution maps using forestry field survey samples from the target year. The quantitative evaluation metrics include Overall Accuracy (OA), Kappa coefficient, Producer's Accuracy (PA), User's Accuracy (UA), and F1 score (F1score), calculated as follows:

[0104]

[0105] Where n is the forest type, N is the total number of forest type samples surveyed in 2023, and X ii Represents the elements on the diagonal of the confusion matrix, that is, the total number of correctly classified pixels in each forest type, X i+ and X +i They represent the column sum and row sum of each forest type, that is, the total number of misclassified and incorrectly classified pixels.

[0106] The classification accuracy of the forest type classification map of the monitoring area in 2023, which was inferred using the method of this embodiment, was compared with the field forestry survey sample point data in 2023. The results are as follows:

[0107] Table 2 shows that the quality of training samples is positively correlated with classification accuracy. Compared to directly applying all historical samples to the target year's forest type classification (OA: 64.94%, kappa: 0.4059), the dual-temporal imagery change detection method achieved higher classification accuracy for remote sensing classification of mountain forest types (OA: 73.27%, kappa: 0.5290). This phenomenon may be primarily due to the prevalence of natural forest growth and succession, which results in a certain proportion of historical samples having transitioned from forest to non-forest by the target year, i.e., false positive samples. The dual-temporal imagery change detection algorithm effectively filters out these samples, improving the quality of training samples and the accuracy of forest type classification to a certain extent. However, the dual-temporal imagery change detection algorithm only detects samples with significant spectral signature changes, ignoring the impact of spectral signature changes caused by forest type growth and succession on sample migration quality. Consequently, the migrated forest type samples still contain a large number of false positive samples. Compared to using only a dual-temporal imagery change detection algorithm, the OA and kappa coefficients of the forest type classification results for Yunnan Province generated using STM-CF increased by 10.06% and 0.1855, respectively. Therefore, compared to existing sample migration methods, the proposed STM-CFS can effectively filter out samples where forest type changes occurred between the historical year and the target year (e.g., pure forest to mixed forest), improving the quality of sample migration and enhancing the classification accuracy of mountain forest types.

[0108] Table 2 Comparison of forest type classification accuracy of different sample migration methods

[0109]

[0110] from Figure 3-Figure 6 It can be seen that the effectiveness of the sample migration method proposed in this embodiment in remote sensing classification of mountain forest types is generally better than that of existing sample migration methods, but different sample migration methods have significant differences in the classification accuracy of each forest type; for example, the classification accuracy of coniferous forests and broad-leaved forests is significantly higher than that of mixed coniferous and broad-leaved bamboo forests. Compared with the use of all historical samples, the classification accuracy of the dual-phase image change detection algorithm for coniferous forests, broad-leaved forests, and mixed coniferous and broad-leaved forests shows a significant upward trend, but the classification accuracy of bamboo forests decreases slightly; the PA of coniferous forests (such as Figure 3 As shown), UA (as Figure 4 as shown) and F1scor (as Figure 5The PA, UA and F1score of broad-leaved forest increased by 0.95%, 15.69% and 0.0890 respectively; the PA and F1score parameters of coniferous and broad-leaved mixed forest increased by 11.84% and 0.0018, while UA decreased by 1.73%; the UA of bamboo forest increased by 8.11%, while PA and F1score decreased by 24.00% and 0.0037. Compared with the dual-temporal change detection method, the STM-CFS sample migration method proposed in this paper significantly improves the classification accuracy of all forest types; for example, the PA, UA and F1score of coniferous forests increased by 12.05%, 6.29% and 0.0963 respectively, the PA, UA and F1score of broad-leaved forests increased by 7.87%, 5.34% and 0.0686 respectively, the PA, UA and F1score of mixed coniferous and broad-leaved forests increased by 29.40%, 37.37% and 0.3810 respectively, and the PA, UA and F1score of bamboo forests increased by 12.10%, 63.19% and 0.6478 respectively. The above research results show that compared with directly applying all historical samples to the forest type classification of the target year, the use of dual-temporal image change detection method to extract unchanged samples effectively increases the quality of training samples and improves the classification accuracy of coniferous forests, broad-leaved forests and mixed coniferous and broad-leaved forests; however, due to the small number of bamboo forest samples (such as Figure 6 (as shown in the figure), and the spectral signature differences between this method and other forest types are small. This method's bamboo forest samples contain a certain proportion of false positive samples indicating changes in the dominant tree species, leading to a significant increase in bamboo forest omission errors. The STM-CFS method effectively retains samples where no forest type change has occurred, improving the quality of training samples and enhancing the classification accuracy of each forest type.

[0111] It can be seen from the above technical solutions that the method for automatic migration of historical forestry field survey samples based on remote sensing forest classification provided by the embodiment of the present invention can effectively use forestry field survey samples of known historical years to infer the remote sensing classification map of forest types in the target year. Case studies in the monitoring area have shown that the method provided by the present invention can effectively improve the migration quality of historical forestry field survey samples, increase the number of effective samples in the target year, and improve the classification accuracy of mountain forest types under small sample conditions of machine learning algorithms. Compared with existing historical sample migration methods, the sample quality of the migration of the present invention is higher, and the results of this embodiment have a higher consistency with the forestry field survey results of the target year, indicating that the method provided by the present invention is more effective for the classification of forest types in a complex terrain area such as the monitoring area.

[0112] The above description is only a preferred embodiment of the present invention and an explanation of the technical principles used. It is not intended to limit the scope of the invention to be protected, but merely represents a preferred embodiment of the present invention. Those skilled in the art should understand that the scope of the invention involved in the present invention is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the inventive concept. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work shall fall within the scope of protection of the present invention.

Claims

1. A method for migrating historical forestry field survey samples based on forest remote sensing classification, characterized in that: The method utilizes the growth succession information of historical samples from the historical year to the target year to migrate historical samples in the same monitoring area where no forest type replacement has occurred to the target year. The method specifically includes the following steps: Step S1, determining the monitoring area, historical years and target years, obtaining historical sample points of forestry field surveys in historical years and target sample points of forestry field surveys in the target year within the monitoring area; Step S2, obtaining long-term multi-source remote sensing data of the historical sample points from the historical year to the target year; Step S3, using long-term multi-source remote sensing data to characterize the forest growth succession characteristics of the historical sample points; the forest growth succession characteristics include spectral similarity of the historical sample points in the historical year and the target year, the temporal trajectory of the historical sample points from the historical year to the target year, and the environmental similarity between the historical sample points and the target sample points of the same type; When calculating the environmental similarity between historical sample points and target sample points of the same type, first, multi-source remote sensing data of the target year is used to characterize the environmental conditions related to the spatial distribution of forest types, including forest canopy characteristics and geographical environment; second, the Gaussian similarity function is used to compare the similarity of environmental conditions between historical sample points and target sample points, which specifically includes the following steps: Step S331, using multi-source remote sensing data acquired in the target year to comprehensively characterize the environmental conditions of each geographical location; Step S332: Calculate the environmental similarity between the historical sample point and the target sample point. The calculation formula is as follows: In formula (4), and are the values ​​of the historical sample point i and the target sample point j on the vth environmental covariate; m is the number of geographical environmental covariates used to characterize the spatial distribution of forest types, which is based on the forest type samples in the target year; L v (·) is used to calculate the geographical environment similarity function between the historical sample point i and the target sample point j on a single environmental variable v, and is calculated using the Gaussian similarity function. The calculation formula is as follows: In formula (5), is the standard deviation of the vth environmental covariate in the entire region, It is the square root of the average deviation of the j-th target sample point on all historical sample points i of the same type with i = 1, 2, ..., k on the v-th environmental covariate: In formula (6), It is a function used to synthesize the similarities of m environmental covariates. The similarities of each environmental covariate are synthesized by using the characteristic importance of each environmental covariate to the forest type classification to obtain the comprehensive environmental similarity between the historical sample point i and the target sample point j: In formula (7), The average reduction precision method in the random forest algorithm was used to evaluate the feature importance of each environmental covariate for regional forest type classification, and features with feature importance greater than 0.00001 were selected to characterize the geographical environment configuration corresponding to each sample point; Step S4, based on the forest growth succession characteristics, determine whether the forest type of the historical sample point has changed from the historical year to the target year; if the three characteristics of the historical sample point in the historical year and the target year are similar, the spectral similarity, the time series trajectory of the historical year and the target year, and the geographical environment similarity of the historical sample point and the sample point of the same type in the target year are similar, then the forest type to which the historical sample point belongs has not changed; otherwise, it is determined that a change has occurred; the threshold segmentation method is used to determine whether the forest type of the historical sample point has changed from the historical year to the target year, and the formula of the threshold segmentation method is as follows: Moreover, Environmental similarity = max(S i1 , S i2 , …, S ij ) In formula (8), Foresttype is the forest type label of the historical sample point; SAD and ED are spectral similarity indicators that quantify the historical sample point between the historical year and the target year, σ1 and σ2 are the thresholds of SAD and ED respectively; disturbance year is the forest disturbance detection result of the historical sample point from the historical year to the target year, and Historical year is the historical year; Environmental similarity is the environmental similarity between the historical sample point and the target sample point of the same type, σ3 is the segmentation threshold of environmental similarity, and S i,j is the environmental similarity between all sample points of the same forest type in the historical sample point i and the target sample point j; Step S5: Migrate forestry field survey sample data of historical sample points that have not been replaced to the target year, and delete historical sample points that have been replaced.

2. The method for migrating historical forestry field survey samples based on forest remote sensing classification according to claim 1 is characterized in that: Step S1 is to obtain forestry field survey sample point data for historical years and target years in the monitoring area. First, the forestry field survey data for historical years and target years in the monitoring area are obtained, and then the surface-to-point method is used to generate historical sample point data and target sample point data maps respectively.

3. The method for migrating historical forestry field survey samples based on forest remote sensing classification according to claim 1 is characterized in that: The long-term multi-source remote sensing data described in step S2 include all available Landsat Collection 2Tier 1 surface reflectance data images, Sentinel-1 / 2 data, SRTM-DEM data, climate data, and soil moisture data on the Google Earth Engine (GEE) platform corresponding to the period from the historical year to the target year.

4. The method for migrating historical forestry field survey samples based on forest remote sensing classification according to claim 1 is characterized in that: The value of σ1 is 0.95; the value of σ2 is 0.

1.

5. The method for migrating historical forestry field survey samples based on forest remote sensing classification according to claim 1 is characterized in that: The value of σ3 is determined according to actual conditions.

6. The method for migrating historical forestry field survey samples based on forest remote sensing classification according to claim 1, characterized in that: The migration model of the forestry field survey sample data of the historical sample points in step S5 is: In formula (9), forest type 历史样本点 For each historical forestry field survey sample, the forest type of the historical sample point from the historical year to the target year is taken. If the forest type has not been replaced at the historical sample point, it means that the forest type label of the sample point has not changed and it is migrated to the target year for forest type remote sensing classification. Otherwise, the type label information of the sample point has been replaced and the sample point is deleted.

Images