A method and system for evaluating the suitability of agricultural machinery operation efficiency at the field level based on remote sensing imagery

Abstract

This invention proposes a method for evaluating the suitability of agricultural machinery operation efficiency within a field based on remote sensing imagery, comprising: S1, extracting detailed field attribute information based on high-resolution remote sensing imagery; S2, extracting field obstacle information and field topography information based on the field attribute information; S3, constructing a set of key evaluation indicators affecting agricultural machinery operation efficiency within the field; S4, assigning weights to the indicators based on a geographic weighted attention mechanism; S5, establishing a suitability evaluation model for agricultural machinery operation efficiency within the field and locating key negative impact indicators; and S6, classifying the suitability level of agricultural machinery operation efficiency based on the natural breakpoint method. This invention also proposes a corresponding system for evaluating the suitability of agricultural machinery operation efficiency within a field based on remote sensing imagery. By constructing a set of key indicators at the field level and innovatively combining geographic weighted regression and attention mechanisms, this invention fully considers regional differences and field characteristics, significantly improving the scientific rigor and practicality of the evaluation.

Description

Technical Field

[0001] This invention relates to the field of farmland monitoring technology, and more specifically, to a method and system for evaluating the suitability of agricultural machinery operating efficiency at the field level based on remote sensing imagery. Background Technology

[0002] A scientific and precise evaluation of the suitability of agricultural machinery operating efficiency at the field level can not only provide data support for site selection in the early stage and evaluation in the later stage of mechanization transformation, but also promote the deep integration of farmland, agricultural machinery and agronomy, and drive the intelligent development of agricultural mechanization. In recent years, the combination of high-resolution remote sensing technology and deep learning has provided a rich data foundation for the evaluation of the suitability of agricultural machinery operating efficiency at the field level. However, a scientific and reasonable evaluation model is also the key to the evaluation of the suitability of agricultural machinery operating efficiency at the field level. Most existing evaluation methods and technologies adopt fixed weighting methods. These methods ignore the combination of macro-regional characteristics and micro-field characteristics, and therefore cannot achieve adaptive and accurate evaluation on a large scale. Most research methods lack sufficient consideration of regional spatial differences, fail to make full use of the dynamic relationship between geographical information and field characteristics, and the research results lack regional specificity; moreover, previous studies have only focused on the level of agricultural machinery operating efficiency suitability at the field level, failing to identify key negative impact factors and lacking specific guidance to transform abstract scores into specific transformation goals. Summary of the Invention

[0003] To address the problems in the background technology, this invention proposes a field-level agricultural machinery driving efficiency suitability evaluation method that combines a large visual segmentation model, a multi-task learning network, geographic weighted regression, and an attention mechanism.

[0004] The present invention provides a method for evaluating the suitability of agricultural machinery operation efficiency within a field based on remote sensing imagery, comprising: S1, extracting detailed field attribute information based on high-resolution remote sensing imagery; S2, extracting field obstacle information and field topography information based on the field attribute information; S3, constructing a set of key evaluation indicators affecting agricultural machinery operation efficiency within the field; S4, assigning weights to the indicators based on a geographic weighted attention mechanism; S5, establishing a suitability evaluation model for agricultural machinery operation efficiency within the field and identifying key negative impact indicators; and S6, classifying the suitability level of agricultural machinery operation efficiency based on the natural breakpoint method.

[0005] The present invention also proposes an evaluation system for the suitability of agricultural machinery operating efficiency at the field level, which includes a computer-executable program capable of performing the steps described above.

[0006] This invention addresses the challenge of achieving adaptive and accurate evaluation of agricultural machinery operating efficiency across large-scale fields by proposing an intelligent evaluation method that combines refinement with broad applicability. This method constructs a set of key indicators at the field level and innovatively combines geographically weighted regression and attention mechanisms, fully considering regional differences and field characteristics, thus significantly improving the scientific rigor and practicality of the evaluation. By integrating remote sensing data, adaptively and dynamically adjusting indicator weights, and precisely locating key influencing factors, this method provides reliable decision support for improving agricultural machinery operation efficiency and precision farmland improvement. Attached Figure Description

[0007] To facilitate understanding of the invention, it will be described in more detail with reference to the specific embodiments shown in the accompanying drawings. These drawings depict only typical embodiments of the invention and should not be considered as limiting the scope of protection of the invention.

[0008] Figure 1 This is a flowchart of the method of the present invention.

[0009] Figure 2 This is a diagram showing the results of vector extraction of the field plots.

[0010] Figure 3 This is a diagram showing the results of obstacle extraction within the field.

[0011] Figure 4 This is a topographic map of the field.

[0012] Figure 5 This is a set of evaluation indicators that are key factors affecting the efficiency of agricultural machinery operation within a field.

[0013] Figure 6 Suitable classification map for agricultural machinery operation within the field. Detailed Implementation

[0014] The embodiments of the present invention are described below with reference to the accompanying drawings to enable those skilled in the art to better understand and implement the present invention. However, the listed embodiments are not intended to limit the present invention. In the absence of conflict, the following embodiments and the technical features in the embodiments can be combined with each other, wherein the same components are indicated by the same reference numerals.

[0015] The method of the present invention includes: S1, extracting detailed field attribute information based on high-resolution remote sensing images; S2, extracting field obstacle information and field topography information based on field boundary information; S3, constructing a set of key evaluation indicators affecting the efficiency of agricultural machinery operation within the field; S4, assigning weights to the indicators based on a geographic weighted attention mechanism; S5, establishing a suitability evaluation model for agricultural machinery operation efficiency within the field and locating key negative impact indicators; S6, classifying the suitability level of agricultural machinery operation efficiency based on the natural breakpoint method.

[0016] In one implementation, in step S1, based on high-resolution remote sensing imagery of the target area, field attribute information is obtained using a deep learning-based land parcel extraction algorithm. In one implementation, step S1 includes S11-S16.

[0017] Preferably, high spatial resolution remote sensing images of the target area without clouds are selected and preprocessed (radiometric calibration, atmospheric correction, etc.).

[0018] S11, perform overlap screening on the masks generated after high-resolution remote sensing image segmentation (preferably SAM segmentation). Based on the stability score as the evaluation criterion, sort the effective mask list of SAM segmentation in descending order. The overlap of masks in the effective mask list is then evaluated sequentially. Masks without overlap can be added to a separate field mask list. Masks with overlap are judged based on indicators such as overlap ratio, number of overlapping masks, and mask quality. Masks meeting the above field mask judgment criteria are retained; otherwise, they are discarded. The stability score is one of the SAM mask quality indicators, and its formula is as follows:

[0019]

[0020] Preferably, in S12, a dilatational erosion operation is performed on the mask after one segmentation to remove linear non-cultivated land features such as small roads and field ridges from the original image. The remaining unsegmented image regions are then used as input for the next segmentation.

[0021] Preferably, in step S13, iterative segmentation is performed. The high-resolution image is initially segmented using a rule-guided dot matrix of specified density to generate an initial effective mask. Based on the initial effective mask, a dilated convolution operation is performed to extract complex regions not covered by the mask, which are then used as iterative images for iterative segmentation. In each round of iterative segmentation, an overlap filtering mechanism is applied to the effective mask. If the number of iterations is greater than 10 (set as needed), segmentation stops. If the number of iterations is less than 10, the segmentation result is compared with the previous result to determine if they are the same. If they are the same, segmentation stops; if they are different, the mask coverage rate is calculated. If the mask coverage rate is less than 90%, iterative segmentation continues; if the coverage rate is greater than or equal to 90%, segmentation stops.

[0022] Preferably, in step S14, the predicted intersection-union ratio (IUU) mask is stitched together: 1) The resulting field grid mask is vectorized to separate and independently number each field mask; 2) The predicted IUU information generated during the segmentation process is transferred to the attributes of the vector elements, and all elements are sorted (in descending order) based on this information; 3) All elements are traversed, and topological relationships are checked and added to the result list in turn. If the current element has a topological relationship with an existing element in the list, it is removed, thereby forming the field area vector segmentation result.

[0023] S15, visualize the final vector results of the fields (see...) Figure 2 ).

[0024] S16, use the rotating caliper algorithm to obtain the circumscribed rectangle vector information of the field.

[0025] In step S2, based on the boundary information of the fields, obstacles (such as trees, utility poles, etc.) are extracted from the fields using a multi-task learning algorithm, and terrain features (such as slope) are obtained by combining them with a digital elevation model, thus achieving a comprehensive analysis of the field attributes. In one embodiment, step S2 includes S21-S26.

[0026] S21, Annotated Dataset: Collect remote sensing image datasets containing field obstacles and perform precise annotation work to generate annotation files.

[0027] S22: Establish a multi-task model, using ResNet as the backbone network to extract shared low-level features from the images. Different specific branches are set for different types of field obstacles: points, lines, and surfaces.

[0028] S23, Feature Fusion and Multi-Task Model Learning: A backbone network is used to extract multi-scale, multi-level features, which are then fused through a feature fusion network. An attention mechanism is then employed to weight the fused features, enhancing the weights of key features and improving overall detection performance.

[0029] S24, Obstacle detection: The multi-task model described above is applied to the remote sensing image of the target area to process different types of obstacles in parallel, and to obtain obstacle information within the field of the target area, which is then stored in Shapefile format.

[0030] S25. Use GIS tools in conjunction with digital elevation models to obtain the topographic features of the fields (such as slope and elevation).

[0031] S26. Use GIS tools to spatially overlay obstacle detection results and terrain feature data with field vectors.

[0032] In step S3, focusing on the driving characteristics of agricultural machinery in the field, and taking into account the field shape, obstacle distribution and terrain, a set of key evaluation indicators that affect the driving efficiency of agricultural machinery is constructed to ensure the relevance of the evaluation.

[0033] On the one hand, through demand analysis, combined with the characteristics of agricultural machinery operations and expert opinions, field morphology, obstacle distribution, and terrain features were identified as the core dimensions. On the other hand, during the indicator selection stage, indicators directly related to efficiency and quantifiable were chosen from the candidate indicators, and redundancy was eliminated through correlation analysis to ensure indicator independence. Based on the above principles, field area, rectangularity, and morphology index were selected for field morphology; obstacle area ratio, number, congestion level, and shortest distance from the field boundary were selected for obstacle distribution; and field slope and field elevation variability were selected for terrain. These nine indicators together constitute the key evaluation indicator set affecting agricultural machinery operating efficiency. Dimensionless values ​​for each indicator were calculated based on field attribute information, obstacle attribute information, and terrain attribute information in S1 and S2.

[0034] The resulting set of indicators not only reflects the geometric characteristics of the field, but also takes into account the impact of internal obstacles and terrain on the operation of agricultural machinery within the field. It can evaluate the suitability level of agricultural machinery operation efficiency at the field level from multiple dimensions, ensuring the comprehensiveness and objectivity of the evaluation process.

[0035] Table 1. Set of evaluation indicators that are key factors affecting the efficiency of agricultural machinery operation in the field.

[0036]

[0037] In step S4, an index weighting method based on a geographically weighted attention mechanism is adopted. A geographically weighted regression model is used to analyze the spatial heterogeneity of farmland characteristics across Chinese provinces to generate provincial-level benchmark weights. The attention mechanism is then used to dynamically adjust the weights based on specific field characteristics, achieving dual-scale evaluation index weighting and significantly improving the regional adaptability of the weighting. In one implementation, steps S41-S42 are described.

[0038] S41 uses a geographically weighted regression model to analyze the spatial heterogeneity of farmland characteristics in various provinces of China, generating provincial-level benchmark weights to reflect the differences in the importance of various indicators among different provinces (such as the large flat fields in Heilongjiang and the small mountainous fields in Guizhou).

[0039] Data is collected to obtain farmland characteristic data for each province (i.e., key evaluation indicators affecting agricultural machinery operating efficiency within fields). This data forms a matrix X, where each row represents farmland characteristic data for a province, and each column corresponds to a farmland characteristic indicator. Historical data on agricultural machinery operating efficiency is also obtained as a reference, denoted as vector y. The elements of vector y correspond to the rows of matrix X, representing the agricultural machinery operating efficiency yi of the i-th province.i Corresponding to its farmland characteristic data X i Assuming there are n provinces and m farmland characteristic indicators, then X is an n×m matrix and y is an n×1 vector.

[0040] (1) Construct a geographically weighted regression (GWR) model, using agricultural machinery operating efficiency as the response variable and farmland characteristic index set as the independent variables. For the i-th province, the model form is as follows:

[0041]

[0042] (2) Wherein, (u i ,v i ) represents the geographical coordinates of the i-th province; β0(u i ,v i ) is the intercept term; β ik (u i ,v i ) is the regression coefficient of the k-th independent variable (i.e., the k-th farmland characteristic indicator) in the i-th province; x ik It is the k-th farmland characteristic index value of the i-th province; ∈ i This is the error term. By fitting this model, the regression coefficient β for each farmland characteristic indicator in each province can be obtained. ik The regression coefficients of all provinces form a coefficient matrix β, which has a size of n×m.

[0043] (3) Calculate the provincial-level benchmark weights. Normalize the regression coefficients of each province to generate the provincial-level benchmark weights. For the k-th indicator of the i-th province, its benchmark weight is... The calculation formula is:

[0044]

[0045] Where, |β ik | represents the absolute value of the regression coefficient, and the denominator represents the sum of the absolute values ​​of the regression coefficients of all indicators for the i-th province. This is how it is calculated. This reflects the importance of the k-th indicator relative to other indicators in the i-th province, and realizes the generation of benchmark weights that take into account the spatial heterogeneity of farmland characteristics in provinces.

[0046] Through the above steps, the spatial heterogeneity of farmland characteristics in various provinces of China was analyzed using a geographically weighted regression model, and provincial-level benchmark weights that reflect the differences in the importance of each indicator to different provinces were generated.

[0047] S42, based on an attention mechanism, calculates attention scores according to the specific characteristics of fields and obstacles (such as area size and slope), dynamically adjusts weights, and finally merges them into the final weights of all indicators at the field level. This dual-scale weighting method helps improve the regional adaptability of weights, ensuring that the evaluation results are consistent with both the overall characteristics of the province and the individual characteristics of the fields. For example, in the Northeast Plain, field area has a relatively high weight, while in the hilly areas of the South, slope and obstacle distribution have more prominent weights.

[0048] (1) Data preparation: Collect characteristic data of specific fields within the province to construct the feature value vector x. t The dimension is 1×m.

[0049] (2) Indicator standardization: The Z-score standardization method is used to standardize the characteristic index values ​​of the field.

[0050]

[0051] Where, μ i It is the mean of the i-th indicator of all fields in the province, σ i It is the standard deviation of the i-th indicator among all fields in the province. The standardized eigenvector of the characteristic indicator values ​​of the field is:

[0052] (3) Calculate the attention score. Utilize the attention mechanism to calculate the attention score for each indicator and perform weighted fusion. For the i-th indicator, its attention score a it See Formula 5 for the calculation to obtain the attention score vector A. t The dimension is 1×m. The fusion formula is Formula 6.

[0053]

[0054]

[0055] Here, α is the fusion coefficient, with a value ranging from [0,1]. The optimal value needs to be determined through methods such as sensitivity analysis. Finally, the final weight vector W at the field level is obtained. t The dimension is 1×m.

[0056] In step S5, based on the weighted indicators, a suitability score for agricultural machinery operating efficiency within the field is calculated using a weighted summation method. Furthermore, contribution analysis is used to identify key negative impact indicators, providing a reference for specific field improvement measures. In one implementation, step S5 includes steps S51-S54.

[0057] S51, Data Preparation: Obtain the field-level weights w from step S4. tj Standardized index values ​​of the field Verify the data to ensure that it is consistent and accurate in terms of format and statistical range, so as to provide a reliable basis for subsequent calculations.

[0058] S52, calculate and statistically analyze the suitability score of agricultural machinery operating efficiency within the field. The weighted summation method is used to calculate the suitability score of the field, and the score results are statistically analyzed.

[0059] (1) Calculate the suitability score. For a specific field in a province, based on the field-level weights and standardized index values ​​obtained in step S4, calculate its agricultural machinery driving efficiency suitability score S using the weighted summation method. t The formula is as follows.

[0060]

[0061] The w here tj It is the weight of the indicator j of the t-th field within the province obtained after fusion in step S4. This is the standardized value of indicator j for the t-th field within the province. Using this formula, a suitability score is calculated for each field within the province, resulting in a score vector S. t Then, using this method, the score vector for each field within the province is calculated.

[0062] (2) Statistical analysis: Perform statistical analysis on the calculated suitability scores. Calculate the mean score. The formula used to reflect the average level of agricultural machinery operating efficiency across all fields in a province is as follows.

[0063]

[0064] Calculate the standard deviation σ S The following formula measures the dispersion of the scores.

[0065]

[0066] S53 identifies key negative impact indicators, primarily by comparing the contribution values ​​of various indicators to screen out the negative factors that have the greatest impact on efficiency.

[0067] (1) Calculate the contribution of the indicator. For the t-th field in the province, calculate the contribution C of indicator j. tj The formula is as follows.

[0068]

[0069] in, Let j be the optimal value of indicator j among all fields in the province (this value needs to be determined based on the actual properties of the indicator; for example, for the slope indicator, a smaller value is optimal; for the field area indicator, a larger value is optimal, etc.). The contribution of each indicator for each field in the province is calculated using this formula, resulting in the contribution matrix C.

[0070] (2) Screen key negative impact indicators and compare the contribution C of each indicator in the t-th field within the province. tj Size, find the indicator with the greatest contribution, and record it as Then, the indicators that appear most frequently in each field in the province are statistically analyzed and identified as the key negative factors that have the greatest impact on the efficiency of agricultural machinery operation in that province.

[0071] S54, combining the spatial distribution characteristics of fields within a province, maps the identified key negative indicators to the specific locations of fields within that province. Using Geographic Information System (GIS) technology, the locations of key negative indicators are marked on the field maps of each province with different colors and symbols, creating a visualization. This visually demonstrates the spatial distribution of key negative indicators across different fields within the province, providing a clear and intuitive reference for subsequently developing targeted agricultural machinery operation optimization strategies.

[0072] In one implementation, in step S6, the suitability score is divided into four levels: excellent, good, medium, and poor using the natural breakpoint method, providing a scientific reference for specific field transformation, including location and quantifiable data.

[0073] The results calculated by the above evaluation model are between 0 and 1. Based on the calculated scores, the evaluation results can be segmented using the natural breakpoint method.

[0074] S61, Data Preparation: Obtain the agricultural machinery driving efficiency suitability score vector S for each field within the province, calculated in step S5. t The suitability score of the t-th plot within the province.

[0075] S62, Natural Breakpoint Method for Grading: The natural breakpoint method clusters data by iterating through the data to maximize between-group variance and minimize within-group variance, while also considering that the range of each group and the elements within each group should be as similar as possible. First, obtain the average suitability score of the fields within the province. The variance SDAM is calculated; next, the range combinations are iterated and the sum of squared deviations of the smallest class mean SDCM is calculated and found; finally, the goodness of fit GVF is calculated and judged. The goodness of fit of variance ranges from 0 to 1, where 1 indicates a very good fit and 0 indicates a very poor fit. The relevant calculation formula is as follows.

[0076]

[0077] Where t is the number of fields within the province; Si The suitability score for the i-th field; k is the number of field category traversal groups within the province; t j S represents the number of plots in group j; ij The suitability score for the i-th field in class j; The average suitability score for field type j.

[0078] The natural breakpoint method described above is used to group similar values, and natural breakpoints are set at the locations with the greatest numerical differences. Taking the experimental area as an example, the natural breakpoint method divides the appropriate score of agricultural machinery driving efficiency within the field into four intervals, i.e., let k=4, with scores from low to high corresponding to four levels: poor, medium, good, and excellent.

[0079] S63, perform field classification mapping, add the classification results of the above fields to the attribute table of the field vector layer using ArcGIS, and visualize the field classification using the unique value method in the symbol system (see...). Figure 6 Based on the grading results, targeted optimization strategies for agricultural machinery operations are developed for fields of different grades. For fields of excellent grade, efficient agricultural machinery operation modes can be further promoted to improve agricultural production efficiency; for fields of poor grade, key negative impact indicators need to be analyzed in detail, and improvement measures should be taken to enhance the suitability of agricultural machinery operation efficiency.

[0080] The present invention also proposes an evaluation system applicable to the classification of the suitability of agricultural machinery operation efficiency in fields, which includes a computer-executable program capable of performing the steps described above.

[0081] Using the obtained field and obstacle vector data, various evaluation indicators related to the suitability of agricultural machinery operation efficiency within the field can be extracted. These indicators are selected based on: field morphology suitability (field area, rectangularity, and morphology index); obstacle distribution suitability (obstacle area ratio, number, congestion level, and shortest distance from field boundary); and field terrain suitability (field slope and field elevation variability). These serve as important indicators for evaluating the suitability of agricultural machinery operation efficiency within the field (Table 1).

[0082] This invention addresses the need for precise evaluation of agricultural machinery operating efficiency within agricultural fields by proposing improved evaluation indicators and weighting methods. Utilizing vector information of fields, field obstacles, and terrain, combined with the requirements of agricultural machinery operations, this invention selects key evaluation indicators based on field morphology, field obstacles, and terrain. A geographically weighted attention-based indicator weighting method is employed to obtain the weights of each indicator, ultimately constructing an evaluation model for the suitability of agricultural machinery operating efficiency within a field, and classifying fields accordingly. This innovative method can be widely applied to modern agricultural machinery operations, promoting the development of intelligent and precise agricultural production methods, and further improving agricultural production efficiency and sustainable development.

[0083] The embodiments described above are merely preferred embodiments of the present invention. The terms "in one embodiment," "in another embodiment," "in yet another embodiment," or "in still another embodiment" used in this specification all refer to one or more of the same or different embodiments according to this disclosure. Ordinary variations and substitutions made by those skilled in the art within the scope of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating the suitability of agricultural machinery operation efficiency within a field based on remote sensing imagery, comprising: S1, extracting detailed field attribute information based on high-resolution remote sensing images; S2, extract field obstacle information and field topography information based on field attribute information; S3, Construct a set of key evaluation indicators that affect the efficiency of agricultural machinery operation within the field; S4, indicator weighting based on a geographic weighted attention mechanism; S5. Establish a suitability evaluation model for agricultural machinery operation efficiency within the field and identify key negative impact indicators. S6, classifying the suitability level of agricultural machinery operating efficiency based on the natural breakpoint method; Step S1 includes: S11, Overlap screening of the mask generated after high-resolution remote sensing image segmentation; S12, perform dilation and erosion operation, perform dilation and erosion operation on the mask after one segmentation, remove linear non-cultivated land features from the original image, and use the remaining unsegmented image regions as input for the next segmentation. S13, perform iterative segmentation, use a rule-cued dot mesh of specified density to perform initial instance segmentation on the high-resolution image, and generate an initial effective mask; S14, perform mask stitching for predicting the cross-union ratio, and vectorize the resulting field grid mask results obtained from the segmentation; S15, visualize the final vector results of the fields; S16, use the rotating caliper algorithm to obtain the bounding rectangle vector information of the field. Step S2 includes: S21, Annotated Dataset: Collect and annotate remote sensing image datasets containing field obstacles; S22, Establish a multi-task model: Use ResNet as the backbone network to extract shared low-level features of the image; S23, Feature Fusion and Multi-Task Model Learning: Use the backbone network to extract multi-scale and multi-level features, and fuse them through a feature fusion network; S24, Obstacle detection: The multi-task model is applied to the remote sensing image of the target area to process different types of obstacles in parallel and to obtain obstacle information in the field of the target area. S25, Use GIS tools in conjunction with digital elevation models to obtain the topographic features of the fields; S26. Use GIS tools to spatially overlay obstacle detection results and terrain feature data with field vectors. In step S4, an index weighting method based on a geographic weighted attention mechanism is adopted. The spatial heterogeneity of farmland characteristics in various provinces of China is analyzed through a geographic weighted regression model to generate provincial benchmark weights. The weights are then dynamically adjusted according to the specific field characteristics using the attention mechanism to achieve dual-scale evaluation index weighting.

2. The method according to claim 1, characterized in that, Step S14 includes: 1) Vectorize the resulting field grid mask from the segmentation to separate and independently number each field mask; 2) Transfer the predicted intersection-union ratio information generated during the segmentation process to the attributes of the vector features, and sort all features based on this information; 3) Traverse all elements, check their topological relationships in turn, and add them to the results list. If the current element has a topological relationship with an existing element in the list, remove it to form the field area vector segmentation result.

3. The method according to claim 1, characterized in that, In step S3, the evaluation index set includes: Evaluation indicators for field morphology include: field area, rectangularity, and morphology index; Evaluation indicators for obstacle distribution include: area percentage, number, crowding level, and shortest distance from field boundary; Evaluation indicators for terrain include: field slope and field elevation variability.

4. The method according to claim 1, characterized in that, Step S4 includes: S41 uses a geographically weighted regression model to analyze the spatial heterogeneity of farmland characteristics in Chinese provinces, generating provincial-level benchmark weights to reflect the differences in the importance of various indicators among different provinces. S42, based on the attention mechanism, calculates attention scores according to the feature values ​​of specific fields and obstacles, dynamically adjusts the weights, and finally merges them into the final weights of all indicators at the field level.

5. The method according to claim 4, characterized in that, Step S5 includes: S51, Data preparation: Obtain the field-level weights and standardized index values ​​of the fields from step S4. S52, calculate and statistically analyze the suitability score of agricultural machinery driving efficiency within the field; S53 identifies key negative impact indicators and filters out the negative factors that have the greatest impact on efficiency by comparing the contribution values ​​of each indicator. S54, combining the spatial distribution characteristics of fields within the province, maps the identified key negative indicators to the location of specific fields within the province.

6. The method according to claim 5, characterized in that, Step S6 includes: S61, Data preparation: Obtain the agricultural machinery driving efficiency suitability score vector for each field in the province calculated in step S5. S62 uses the natural breakpoint method to classify levels; S63, perform field grading mapping, add the grading results of the fields to the attribute table of the field vector layer, and use the unique value method in the symbol system to visualize the field grading.

7. A field-level agricultural machinery driving efficiency suitability evaluation system based on remote sensing imagery, characterized in that, It includes computer-executable code, which, when executed, implements the method as described in any one of claims 1-6.

Images