Grassland degradation grade diagnosis and restoration system based on AI image recognition

Abstract

The invention relates to the technical field of grassland degradation restoration, and discloses a grassland degradation grade diagnosis and restoration system based on AI image recognition, and the system comprises an acquisition module which is configured to acquire image data of a to-be-monitored grassland, analyze the image data, and determine a degradation evaluation value of the to-be-monitored grassland based on an analysis result; the judgment module is configured to collect a preset ecological index of the to-be-monitored grassland and judge whether to adjust the degradation evaluation value or not according to the preset ecological index; the adjusting module is configured to determine an adjusting coefficient of the degradation evaluation value based on the rhizome density and obtain a degradation evaluation final value; the grade diagnosis module is configured to determine the degradation grade of the to-be-monitored grassland according to the degradation evaluation final value; and the restoration module is configured to determine a restoration strategy of the to-be-monitored grassland according to the degradation level. By combining the AI image recognition technology, the system can efficiently and accurately diagnose and repair the grassland degradation problem.

Description

Technical Field

[0001] This invention relates to the field of grassland degradation restoration technology, and more specifically, to a grassland degradation level diagnosis and restoration system based on AI image recognition. Background Technology

[0002] With the intensification of global climate change and human activities, grassland degradation has become an increasingly serious problem. Grassland degradation not only leads to reduced vegetation cover and biodiversity, but also triggers a series of ecological and environmental problems such as soil erosion and desertification, posing a serious threat to regional ecological security and socio-economic development. Traditional methods for monitoring and assessing grassland degradation mainly rely on field surveys and sampling analysis. This method is not only costly in terms of manpower, resources, and time, but also difficult to achieve real-time, dynamic monitoring of large areas of grassland. Furthermore, in terms of grassland restoration, past restoration strategies often lack specificity, failing to adopt the most appropriate restoration measures according to different levels of degradation, resulting in poor restoration effects and significant resource waste.

[0003] Therefore, it is necessary to design a grassland degradation level diagnosis and restoration system based on AI image recognition to solve the problems existing in the current technology. Summary of the Invention

[0004] In view of this, the present invention proposes a grassland degradation level diagnosis and restoration system based on AI image recognition, which aims to solve the above problems.

[0005] This invention proposes a grassland degradation level diagnosis and restoration system based on AI image recognition, comprising: The acquisition module is configured to identify the grassland to be monitored, acquire image data of the grassland to be monitored, analyze the image data, and determine the degradation assessment value of the grassland to be monitored based on the analysis results. The judgment module is configured to collect preset ecological indicators of the grassland to be monitored, and determine whether to adjust the degradation assessment value based on the preset ecological indicators; The adjustment module is configured to, when it is determined that the degradation assessment value needs to be adjusted, collect the root and stem density of the grassland to be monitored, determine the adjustment coefficient of the degradation assessment value based on the root and stem density, and obtain the final degradation assessment value. The degradation diagnosis module is configured to determine the degradation level of the grassland to be monitored based on the final value of the degradation assessment. The restoration module is configured to determine a restoration strategy for the grassland to be monitored based on the degradation level.

[0006] Furthermore, when determining the degradation assessment value of the grassland to be monitored based on the analysis results, the following steps are included: The image data is analyzed to obtain the vegetation status characteristic value, grassland bareness characteristic value, and soil erosion characteristic value of the grassland to be monitored; A multi-source image feature vector is constructed based on the vegetation condition feature value, grassland bareness feature value, and soil erosion feature value. The feature vectors of the multi-source images are compared with the historical image set, and the degradation assessment value of the grassland to be monitored is determined based on the comparison results.

[0007] Furthermore, when determining the degradation assessment value of the grassland to be monitored based on the comparison results, the following steps are included: When there is a historical multi-source image feature vector in the historical image set that is the same as the multi-source image feature vector, the historical degradation evaluation value corresponding to the historical multi-source image feature vector is used as the degradation evaluation value; When there is no historical multi-source image feature vector in the historical image set that is the same as the multi-source image feature vector, the degradation assessment baseline value of the grassland to be monitored is determined based on the grassland bareness feature value. The degradation assessment value is obtained by compensating the degradation assessment baseline value based on the vegetation condition characteristic value and the soil erosion characteristic value.

[0008] Furthermore, when determining the baseline value for degradation assessment of the grassland to be monitored based on the grassland bareness characteristic value, the following steps are included: The bare grassland feature value is compared with the first bare grassland feature value and the second bare grassland feature value, and the degradation assessment baseline value is determined based on the comparison result; wherein, the first bare grassland feature value is less than the second bare grassland feature value; When the grassland bareness characteristic value is less than the first grassland bareness characteristic value, the degradation assessment baseline value is determined to be the first degradation assessment baseline value; When the grassland bareness characteristic value is greater than or equal to the first grassland bareness characteristic value and less than the second grassland bareness characteristic value, the degradation assessment baseline value is determined to be the second degradation assessment baseline value. When the bare grassland characteristic value is greater than or equal to the second bare grassland characteristic value, the degradation assessment baseline value is determined to be the third degradation assessment baseline value.

[0009] Furthermore, when obtaining the degradation assessment value by compensating the baseline value of the degradation assessment based on the vegetation condition characteristic value and the soil erosion characteristic value, the following steps are included: The vegetation condition characteristic value is compared with the vegetation condition characteristic threshold, the soil erosion characteristic value is compared with the soil erosion characteristic threshold, and the compensation coefficient of the degradation assessment baseline value is determined based on the comparison results. When the vegetation condition characteristic value is greater than the vegetation condition characteristic threshold and the soil erosion characteristic value is greater than the soil erosion characteristic threshold, the compensation coefficient is determined to be the first compensation coefficient. When the vegetation condition characteristic value is greater than the vegetation condition characteristic threshold and the soil erosion characteristic value is less than or equal to the soil erosion characteristic threshold, the compensation coefficient is determined to be the second compensation coefficient. When the vegetation condition characteristic value is less than or equal to the vegetation condition characteristic threshold, and the soil erosion characteristic value is greater than the soil erosion characteristic threshold, the compensation coefficient is determined to be the third compensation coefficient. When the vegetation condition characteristic value is less than or equal to the vegetation condition characteristic threshold, and the soil erosion characteristic value is less than or equal to the soil erosion characteristic threshold, the compensation coefficient is determined to be the fourth compensation coefficient.

[0010] Furthermore, when determining whether to adjust the degradation assessment value based on the preset ecological indicators, the process includes: Collect real-time ecological indicators of the grassland to be monitored; Calculate the degree of deviation between each real-time ecological indicator and its corresponding preset ecological indicator; The deviation index of the comprehensive ecological index is obtained by normalizing and weighting all the aforementioned deviations. The determination of whether to adjust the degradation assessment value is based on the comprehensive ecological index deviation index; If the deviation index of the comprehensive ecological index is greater than 0.45, it is determined that the degradation assessment value should be adjusted. If the deviation index of the comprehensive ecological index is less than or equal to 0.45, it is determined that the degradation assessment value will not be adjusted.

[0011] Further, when determining the adjustment coefficient for the degradation assessment value based on the rootstock density and obtaining the final degradation assessment value, the process includes: A fusion vector group is constructed based on the root and stem density and the comprehensive ecological index deviation index. Calculate the Euclidean distance between the fused vector group and each historical fused vector group, and extract the minimum Euclidean distance; The adjustment coefficient is determined based on the minimum Euclidean distance.

[0012] Further, determining the adjustment coefficient based on the minimum Euclidean distance includes: When the minimum Euclidean distance is unique, the standard Euclidean distance is obtained, and the difference between the standard Euclidean distance and the minimum Euclidean distance is calculated and denoted as the Euclidean distance difference. The Euclidean distance difference is compared with a preset adjustment coefficient mapping table, and the adjustment coefficient is determined based on the comparison result. When the minimum Euclidean distance is not unique, the average value of all the minimum Euclidean distances is obtained and recorded as the average Euclidean distance. The adjustment coefficient is determined based on the average Euclidean distance. The product of the adjustment coefficient and the degradation assessment value is taken as the final degradation assessment value.

[0013] Further, when determining the degradation level of the grassland to be monitored based on the final degradation assessment value, the following steps are included: The final value of the degradation assessment is compared with the first final value of the degradation assessment and the second final value of the degradation assessment, and the degradation level is determined based on the comparison result; wherein, the first final value of the degradation assessment is less than the second final value of the degradation assessment. When the final value of the degradation assessment is less than or equal to the first final value of the degradation assessment, the degradation level is determined to be mild, and the grassland to be monitored is marked as a mildly degraded area; When the final value of the degradation assessment is greater than the first final value of the degradation assessment and less than or equal to the second final value of the degradation assessment, the degradation level is determined to be moderate, and the grassland to be monitored is marked as a moderately degraded area. When the final value of the degradation assessment is greater than the second final value of the degradation assessment, the degradation level is determined to be severe, and the grassland to be monitored is marked as a severely degraded area.

[0014] Furthermore, when determining the restoration strategy for the grassland to be monitored based on the degradation level, the following steps are included: For the mildly degraded areas, biological bud-promoting technology is implemented, including the application of plant growth promoters and slow-release compound fertilizers; For the moderately degraded areas, root cutting and shoot promotion techniques are implemented, using machinery to cut the sod layer, and water and fertilizer management is carried out simultaneously. For the severely degraded areas, the tiller island planting technique was implemented.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: The grassland degradation level diagnosis and restoration system based on AI image recognition provided by this invention combines AI image recognition technology, enabling efficient and accurate diagnosis and restoration of grassland degradation problems. The system acquires image data of the grassland to be monitored through a data acquisition module and analyzes it to determine degradation assessment values, thereby changing the time-consuming and labor-intensive traditional method that relies on field surveys and sampling analysis, and achieving real-time, dynamic monitoring of large areas of grassland. The judgment module determines whether to adjust the degradation assessment values ​​based on preset ecological indicators, ensuring that the assessment results are closer to the actual situation. When necessary, the adjustment module determines the adjustment coefficient based on root density to obtain the final degradation assessment value, further improving the accuracy of the assessment. The level diagnosis module accurately classifies the grassland degradation level based on the final degradation assessment value, providing a clear basis for subsequent restoration work. The restoration module formulates targeted restoration strategies according to different degradation levels, avoiding the problems of poor results and resource waste caused by the lack of specificity in previous restoration strategies. The application of this system effectively improves the efficiency and accuracy of grassland degradation monitoring and assessment, helping to promptly identify grassland degradation problems and take appropriate restoration measures. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a structural block diagram of a grassland degradation level diagnosis and restoration system based on AI image recognition provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] See Figure 1 As shown in some embodiments of this application, this embodiment provides a grassland degradation level diagnosis and restoration system based on AI image recognition, including: The acquisition module is configured to identify the grassland to be monitored, acquire image data of the grassland to be monitored, analyze the image data, and determine the degradation assessment value of the grassland to be monitored based on the analysis results. The judgment module is configured to collect preset ecological indicators of the grassland to be monitored, and determine whether to adjust the degradation assessment value based on the preset ecological indicators; The adjustment module is configured to, when it is determined that the degradation assessment value needs to be adjusted, collect the root and stem density of the grassland to be monitored, determine the adjustment coefficient of the degradation assessment value based on the root and stem density, and obtain the final degradation assessment value. The degradation diagnosis module is configured to determine the degradation level of the grassland to be monitored based on the final value of the degradation assessment. The restoration module is configured to determine a restoration strategy for the grassland to be monitored based on the degradation level.

[0020] It is understood that the grassland degradation level diagnosis and restoration system based on AI image recognition provided in this embodiment, combined with AI image recognition technology, can efficiently and accurately diagnose and restore grassland degradation problems. The system acquires image data of the grassland to be monitored through a data acquisition module and analyzes it to determine degradation assessment values, thereby changing the traditional time-consuming and labor-intensive method that relies on field surveys and sampling analysis, and realizing real-time, dynamic monitoring of large areas of grassland. The judgment module decides whether to adjust the degradation assessment values ​​based on preset ecological indicators, ensuring that the assessment results are closer to the actual situation. The adjustment module, when necessary, determines the adjustment coefficient based on root density to obtain the final degradation assessment value, further improving the accuracy of the assessment. The level diagnosis module accurately classifies the grassland degradation level based on the final degradation assessment value, providing a clear basis for subsequent restoration work. The restoration module formulates targeted restoration strategies according to different degradation levels, avoiding the problems of poor results and resource waste caused by the lack of specificity in previous restoration strategies. The application of this system effectively improves the efficiency and accuracy of grassland degradation monitoring and assessment, helping to promptly identify grassland degradation problems and take appropriate restoration measures.

[0021] Specifically, when determining the degradation assessment value of the grassland to be monitored based on the analysis results, the following are included: The image data is analyzed to obtain the vegetation status characteristic value, grassland bareness characteristic value, and soil erosion characteristic value of the grassland to be monitored; A multi-source image feature vector is constructed based on the vegetation condition feature value, grassland bareness feature value, and soil erosion feature value. The feature vectors of the multi-source images are compared with the historical image set, and the degradation assessment value of the grassland to be monitored is determined based on the comparison results.

[0022] Understandably, the preferred vegetation condition characteristic values, grassland bareness characteristic values, and soil erosion characteristic values ​​are vegetation cover, grassland bareness area percentage, and gully texture density, respectively. Vegetation cover directly reflects the vigor of vegetation growth in the grassland; higher cover indicates better vegetation condition, while lower cover suggests potential degradation. Grassland bareness area percentage reflects the proportion of the grassland surface without vegetation cover; a larger percentage indicates more severe grassland degradation. Gully texture density reflects the degree of soil erosion; higher density indicates more severe soil erosion and a higher likelihood of grassland degradation.

[0023] Specifically, when determining the degradation assessment value of the grassland to be monitored based on the comparison results, the following are included: When there is a historical multi-source image feature vector in the historical image set that is the same as the multi-source image feature vector, the historical degradation evaluation value corresponding to the historical multi-source image feature vector is used as the degradation evaluation value; When there is no historical multi-source image feature vector in the historical image set that is the same as the multi-source image feature vector, the degradation assessment baseline value of the grassland to be monitored is determined based on the grassland bareness feature value. The degradation assessment value is obtained by compensating the degradation assessment baseline value based on the vegetation condition characteristic value and the soil erosion characteristic value.

[0024] Understandably, the above method can make full use of historical data. When the same feature vector exists, the historical evaluation value is directly used, which improves the evaluation efficiency. When it does not exist, the base value is first determined based on the grassland bareness feature value, and then compensation is made by combining the vegetation condition and soil erosion feature value, so that the evaluation results are more comprehensive and accurate.

[0025] Specifically, when determining the baseline value for degradation assessment of the grassland to be monitored based on the grassland bareness characteristic value, the following are included: The bare grassland feature value is compared with the first bare grassland feature value and the second bare grassland feature value, and the degradation assessment baseline value is determined based on the comparison result; wherein, the first bare grassland feature value is less than the second bare grassland feature value; When the grassland bareness characteristic value is less than the first grassland bareness characteristic value, the degradation assessment baseline value is determined to be the first degradation assessment baseline value; When the grassland bareness characteristic value is greater than or equal to the first grassland bareness characteristic value and less than the second grassland bareness characteristic value, the degradation assessment baseline value is determined to be the second degradation assessment baseline value. When the bare grassland characteristic value is greater than or equal to the second bare grassland characteristic value, the degradation assessment baseline value is determined to be the third degradation assessment baseline value.

[0026] Understandably, the baseline values ​​for degradation assessment follow this order: First baseline value < Second baseline value < Third baseline value. This is because a smaller grassland bareness characteristic value indicates a lower degree of grassland degradation, and therefore a smaller baseline value for degradation assessment; conversely, a larger grassland bareness characteristic value indicates more severe grassland degradation, and therefore a larger baseline value for degradation assessment. This method of determining baseline values ​​for degradation assessment based on grassland bareness characteristic values ​​can scientifically reflect the initial degradation state of grassland.

[0027] Specifically, when compensating the baseline degradation assessment value based on the vegetation condition characteristic value and soil erosion characteristic value to obtain the degradation assessment value, the following steps are included: The vegetation condition characteristic value is compared with the vegetation condition characteristic threshold, the soil erosion characteristic value is compared with the soil erosion characteristic threshold, and the compensation coefficient of the degradation assessment baseline value is determined based on the comparison results. When the vegetation condition characteristic value is greater than the vegetation condition characteristic threshold and the soil erosion characteristic value is greater than the soil erosion characteristic threshold, the compensation coefficient is determined to be the first compensation coefficient. When the vegetation condition characteristic value is greater than the vegetation condition characteristic threshold and the soil erosion characteristic value is less than or equal to the soil erosion characteristic threshold, the compensation coefficient is determined to be the second compensation coefficient. When the vegetation condition characteristic value is less than or equal to the vegetation condition characteristic threshold, and the soil erosion characteristic value is greater than the soil erosion characteristic threshold, the compensation coefficient is determined to be the third compensation coefficient. When the vegetation condition characteristic value is less than or equal to the vegetation condition characteristic threshold, and the soil erosion characteristic value is less than or equal to the soil erosion characteristic threshold, the compensation coefficient is determined to be the fourth compensation coefficient.

[0028] Understandably, the compensation coefficients follow the order: First compensation coefficient < Second compensation coefficient < Third compensation coefficient < Fourth compensation coefficient. When both the vegetation condition characteristic value and the soil erosion characteristic value are greater than the threshold, it indicates that although the grassland vegetation is relatively healthy, soil erosion is severe, and the overall degradation still needs attention, so the compensation coefficient is small. When the vegetation condition characteristic value is greater than the threshold while the soil erosion characteristic value is less than or equal to the threshold, it indicates that the grassland vegetation is growing well and the degree of soil erosion is low, resulting in a relatively low degree of grassland degradation, and the compensation coefficient increases. When the vegetation condition characteristic value is less than or equal to the threshold and the soil erosion characteristic value is greater than the threshold, it means that the grassland vegetation is not growing well and soil erosion is severe, resulting in a high degree of grassland degradation, and the compensation coefficient further increases. When both the vegetation condition characteristic value and the soil erosion characteristic value are less than or equal to the threshold, it indicates that the grassland vegetation is growing poorly and there are also some problems with soil erosion, resulting in the most severe grassland degradation, and the compensation coefficient is the largest. By determining the compensation coefficient based on the comparison of vegetation condition characteristic values ​​and soil erosion characteristic values ​​with their respective thresholds, the baseline degradation assessment value can be adjusted more precisely, thus making the final degradation assessment value more accurately reflect the actual degradation status of the grassland under monitoring. Then, the baseline degradation assessment value is calculated based on the determined compensation coefficient; that is, the degradation assessment value equals the baseline degradation assessment value multiplied by the compensation coefficient, thus completing the final determination of the degradation assessment value.

[0029] Specifically, when determining whether to adjust the degradation assessment value based on the preset ecological indicators, the following steps are included: Collect real-time ecological indicators of the grassland to be monitored; Calculate the degree of deviation between each real-time ecological indicator and its corresponding preset ecological indicator; The deviation index of the comprehensive ecological index is obtained by normalizing and weighting all the aforementioned deviations. The determination of whether to adjust the degradation assessment value is based on the comprehensive ecological index deviation index; If the deviation index of the comprehensive ecological index is greater than 0.45, it is determined that the degradation assessment value should be adjusted. If the deviation index of the comprehensive ecological index is less than or equal to 0.45, it is determined that the degradation assessment value will not be adjusted.

[0030] Understandably, the preferred real-time ecological indicators are soil moisture content, soil temperature, and soil nutrient concentration. Soil moisture content is crucial for grassland vegetation growth; adequate moisture ensures normal physiological activities, while too much or too little water negatively impacts vegetation health and consequently affects the degree of grassland degradation. Soil temperature affects vegetation growth rate and microbial activity; different grassland vegetation types have different tolerance ranges for soil temperature, and exceeding these ranges may lead to poor vegetation growth. Soil nutrient concentration provides the necessary material basis for vegetation growth; insufficient nutrients will cause slow growth, stunted growth, or even death, thus exacerbating grassland degradation.

[0031] Understandably, for soil moisture content, the actual measured value of the grassland to be monitored is first obtained and compared with the preset standard value for that grassland type. The difference between the actual value and the standard value is calculated, and for ease of unified analysis, the difference is divided by the standard value to obtain the relative deviation value. For example, if the preset standard soil moisture content is 30% and the actual measured value is 25%, the difference is -5%, and the relative deviation value is approximately -0.17. For soil temperature, the actual measured value is also obtained and compared with the preset standard value. The difference between the actual value and the standard value is calculated, and then divided by the standard value to obtain the relative deviation value. Assuming the preset standard soil temperature is 20℃ and the actual measured value is 23℃, the difference is 3℃, and the relative deviation value is 0.15. For soil nutrient concentration, calculations are performed separately for major nutrient elements such as nitrogen, phosphorus, and potassium. Taking nitrogen as an example, the actual measured concentration is obtained and compared with the preset standard value, and the relative deviation value is calculated using the above method. The other elements are calculated sequentially. After obtaining the relative deviation values ​​of each indicator, weights are assigned according to the importance of each indicator to grassland degradation. For example, soil moisture content is weighted at 0.4, soil temperature at 0.3, and soil nutrient concentration at 0.3. The relative deviation values ​​of each indicator are multiplied by their corresponding weights and then summed to obtain a comprehensive deviation value. This value is used in subsequent normalized weighted calculations to obtain a comprehensive ecological indicator deviation index. This index ranges from 0 to 1, is dimensionless, and directly reflects the overall deviation between the real-time ecological indicators of the monitored grassland and the preset ecological indicators.

[0032] Understandably, by collecting these real-time ecological indicators and comparing them with preset ecological indicators to calculate the degree of deviation, and then performing normalized weighted calculations to obtain the comprehensive ecological indicator deviation index, the ecological status of the grassland under monitoring can be comprehensively and objectively reflected. When the comprehensive ecological indicator deviation index is greater than 0.45, it indicates that the actual ecological status of the grassland deviates significantly from the preset ideal ecological status. In this case, it is necessary to adjust the degradation assessment value to make the assessment results more consistent with the actual degradation situation of the grassland. When the comprehensive ecological indicator deviation index is less than or equal to 0.45, it indicates that the ecological status of the grassland is relatively close to the preset indicators, and no adjustment to the degradation assessment value is required. This judgment mechanism based on real-time ecological indicators further improves the accuracy and reliability of the system's assessment of grassland degradation, providing a more scientific basis for subsequent restoration work.

[0033] Specifically, when determining the adjustment coefficient for the degradation assessment value based on the rootstock density and obtaining the final degradation assessment value, the process includes: A fusion vector group is constructed based on the root and stem density and the comprehensive ecological index deviation index. Calculate the Euclidean distance between the fused vector group and each historical fused vector group, and extract the minimum Euclidean distance; The adjustment coefficient is determined based on the minimum Euclidean distance.

[0034] Understandably, root and stem density is an important indicator reflecting the growth status of grassland vegetation. It, along with the comprehensive ecological index deviation index, forms a fusion vector set that comprehensively reflects the ecological characteristics and degradation status of the grassland. Historical fusion vector sets are accumulated from previous monitoring and analysis of grasslands with different degrees of degradation, containing rich empirical data. By calculating the Euclidean distance between the current fusion vector set and each historical fusion vector set, the similarity between the current grassland ecological characteristics and historical conditions can be quantified. The historical condition corresponding to the smallest Euclidean distance is closest to the current ecological characteristics of the grassland to be monitored.

[0035] Specifically, determining the adjustment coefficient based on the minimum Euclidean distance includes: When the minimum Euclidean distance is unique, the standard Euclidean distance is obtained, and the difference between the standard Euclidean distance and the minimum Euclidean distance is calculated and denoted as the Euclidean distance difference. The Euclidean distance difference is compared with a preset adjustment coefficient mapping table, and the adjustment coefficient is determined based on the comparison result. When the minimum Euclidean distance is not unique, the average value of all the minimum Euclidean distances is obtained and recorded as the average Euclidean distance. The adjustment coefficient is determined based on the average Euclidean distance. The product of the adjustment coefficient and the degradation assessment value is taken as the final degradation assessment value.

[0036] Understandably, the standard Euclidean distance is a representative value derived from statistical analysis of multiple sets of historical data. It reflects the typical distance between fusion vector groups under ideal ecological conditions. By calculating the difference between the standard Euclidean distance and the minimum Euclidean distance, the degree of deviation between the current grassland ecological characteristics and the ideal condition can be measured. The preset adjustment coefficient mapping table is summarized from extensive experiments and practices, and it maps the range of Euclidean distance differences to the corresponding adjustment coefficients. When the minimum Euclidean distance is unique, comparing the calculated Euclidean distance difference with the mapping table allows for quick and accurate determination of the adjustment coefficient. When the minimum Euclidean distance is not unique, calculating the average of all minimum Euclidean distances is to comprehensively consider multiple relatively close historical situations, making the determined adjustment coefficient more reasonable. Multiplying the adjustment coefficient by the degradation assessment value yields the final degradation assessment value. This allows for correction of the initial degradation assessment value based on the current actual ecological characteristics of the grassland, making the final assessment result more reflective of the true degree of grassland degradation.

[0037] Understandably, determining the adjustment coefficient based on the average Euclidean distance involves the following steps: First, obtain the standard Euclidean distance and calculate the difference between it and the average Euclidean distance, recording this difference as the new Euclidean distance difference. Next, compare this new Euclidean distance difference with a pre-defined adjustment coefficient mapping table. Since the average Euclidean distance incorporates multiple similar historical cases, the correspondence in the mapping table must be carefully considered during the comparison. When the new Euclidean distance difference falls within a specific range in the mapping table, the adjustment coefficient corresponding to that range is determined. This method of determining the adjustment coefficient based on the average Euclidean distance fully considers the combined influence of multiple similar historical cases, avoiding the potential bias of a single minimum Euclidean distance, thus making the determination of the adjustment coefficient more scientific and reasonable.

[0038] Specifically, determining the degradation level of the grassland to be monitored based on the final degradation assessment value includes: The final value of the degradation assessment is compared with the first final value of the degradation assessment and the second final value of the degradation assessment, and the degradation level is determined based on the comparison result; wherein, the first final value of the degradation assessment is less than the second final value of the degradation assessment. When the final value of the degradation assessment is less than or equal to the first final value of the degradation assessment, the degradation level is determined to be mild, and the grassland to be monitored is marked as a mildly degraded area; When the final value of the degradation assessment is greater than the first final value of the degradation assessment and less than or equal to the second final value of the degradation assessment, the degradation level is determined to be moderate, and the grassland to be monitored is marked as a moderately degraded area. When the final value of the degradation assessment is greater than the second final value of the degradation assessment, the degradation level is determined to be severe, and the grassland to be monitored is marked as a severely degraded area.

[0039] Understandably, the first and second final degradation assessment values ​​are representative critical values ​​determined based on extensive grassland degradation data and actual research analysis. They serve as key boundaries for classifying different degradation levels, providing a clear standard for accurately assessing the degree of degradation of the grassland under monitoring. Comparing the final degradation assessment value with these two critical values ​​allows for a rapid and clear determination of the grassland's degradation level. Slightly degraded areas indicate that while the grassland has shown some signs of degradation, its overall ecological condition remains relatively good, and vegetation growth and ecological functions still have significant potential for recovery. Moderately degraded areas show relatively severe degradation, with vegetation growth and ecological functions significantly affected. Severely degraded areas demonstrate severe damage to the grassland ecosystem, with low vegetation cover, deteriorated soil quality, and severely impaired ecological functions. This method of determining degradation levels based on final degradation assessment values ​​enables the development of targeted restoration plans for grasslands with different degrees of degradation, improving the efficiency and effectiveness of grassland restoration efforts, and thus better protecting and restoring grassland ecosystems.

[0040] Specifically, determining the restoration strategy for the grassland to be monitored based on the degradation level includes: For the mildly degraded areas, biological bud-promoting technology is implemented, including the application of plant growth promoters and slow-release compound fertilizers; For the moderately degraded areas, root cutting and shoot promotion techniques are implemented, using machinery to cut the sod layer, and water and fertilizer management is carried out simultaneously. For the severely degraded areas, the tiller island planting technique was implemented.

[0041] Understandably, different restoration strategies are formulated based on the characteristics of grasslands at different degradation levels. In mildly degraded areas, applying plant growth promoters can stimulate plant cell division and growth, enhance stress resistance, and slow-release compound fertilizers can continuously provide nutrients, promoting vegetation restoration and improving the ecology. In moderately degraded areas, the sod layer affects water and nutrient infiltration and root growth. Root pruning to promote shoot growth creates growth space by cutting the sod layer, while simultaneous water and fertilizer management replenishes water and nutrients, promoting shoot growth and increasing vegetation cover and ecological function. In severely degraded areas, the ecosystem is severely damaged. Sprout island planting technology establishes vegetation growth points, expands and improves the surrounding soil environment, attracts seed growth, and restores the ecosystem. Implementing corresponding strategies for different degradation levels can accurately solve problems, improve restoration success rates, and achieve sustainable development. During implementation, continuous monitoring of the restoration situation is necessary, and the plan should be adjusted according to the actual results to ensure the healthy and stable development of grasslands.

[0042] Understandably, the tiller island planting technique specifically includes: (1) Artificial cultivation of tillers: Tillers of the target high-quality grasses are cultivated at high density in a controlled environment; field planting on tillers islands: (2) The cultivated tillers with root ball soil are dug up in clusters to form “tiller islands” and planted in the planting holes prepared in advance in the severely degraded area; (3) Post-planting management: During the critical growth period after planting, necessary water supply and growth maintenance should be carried out.

[0043] Understandably, artificial cultivation of tillers in a controlled environment offers numerous advantages. A controlled environment allows for precise regulation of temperature, humidity, and light, providing the most suitable growth environment for the tillers of the target high-quality grasses, significantly improving the success rate and quality of tiller cultivation. High-density cultivation makes full use of space, cultivating a large number of tillers within a limited area, meeting the needs of large-scale restoration in severely degraded areas. In the field planting stage of the tiller islands, it is crucial to dig up the tillers in clusters with rhizosphere soil. The rhizosphere soil contains beneficial microorganisms symbiotic with the tillers and abundant nutrients. These microorganisms promote the absorption and utilization of nutrients by the tillers, enhancing their ability to adapt to the external environment. The "tiller islands" are then planted in pre-prepared planting holes. The size and depth of the planting holes should be rationally designed according to the size of the tillers and the root growth to ensure stable growth. Post-planting management is also essential. During the critical growth period, necessary water replenishment should be provided. The amount and timing of irrigation should be arranged reasonably according to weather conditions and soil moisture to avoid over-irrigation or under-irrigation, which can adversely affect the growth of tillers. At the same time, growth maintenance also includes timely removal of weeds to prevent them from competing with tillers for nutrients and space; monitoring and controlling pests and diseases to ensure the healthy growth of tillers.

[0044] Understandably, the tiller islands are laid out in the field in a non-uniform clustered distribution, with a planting density sufficient to ensure the formation of continuous vegetation patches during the growing season.

[0045] It is understandable that the plant growth promoters used in biological bud-promoting technology include, but are not limited to, humic acid and seaweed extracts, and the proportion of slow-release compound fertilizer is determined according to the nutritional needs of the target grass species.

[0046] Understandably, in the root-cutting and shoot-promoting technique, the depth and row spacing of the root-cutting operation are adjusted according to the thickness and compactness of the straw layer, and the operation time is selected in the season that is conducive to the recovery and growth of plants.

[0047] It is understandable that a grassland will be divided into several grasslands to be monitored. Then, restoration measures for the slightly degraded, moderately degraded and severely degraded areas of this grassland will be implemented in a coordinated manner over one or more growth cycles. Through the synergistic effect between different technologies, the overall positive succession of the grassland community can be promoted.

[0048] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0049] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. An AI image recognition-based grassland degradation level diagnosis and repair system, characterized in that, The method comprises the following steps: A collection module is configured to determine a to-be-monitored grassland, collect image data of the to-be-monitored grassland, and analyze the image data, and determine a degradation evaluation value of the to-be-monitored grassland based on the analysis result; A judgment module is configured to collect a preset ecological index of the to-be-monitored grassland, and determine whether to adjust the degradation evaluation value according to the preset ecological index; An adjustment module is configured to, when it is determined to adjust the degradation evaluation value, collect a root density of the to-be-monitored grassland, determine an adjustment coefficient of the degradation evaluation value based on the root density, and obtain a final degradation evaluation value; A grade diagnosis module is configured to determine a degradation grade of the to-be-monitored grassland according to the final degradation evaluation value; A repair module is configured to determine a repair strategy of the to-be-monitored grassland according to the degradation grade. 2.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 1, characterized in that, When the degradation evaluation value of the to-be-monitored grassland is determined based on the analysis result, the method comprises the following steps: The image data is analyzed to obtain a vegetation condition characteristic value, a grassland bareness characteristic value, and a soil erosion characteristic value of the to-be-monitored grassland; A multi-source image feature vector is constructed according to the vegetation condition characteristic value, the grassland bareness characteristic value, and the soil erosion characteristic value; The multi-source image feature vector is compared with a historical image set, and the degradation evaluation value of the to-be-monitored grassland is determined according to the comparison result. 3.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 2, characterized in that, When the degradation evaluation value of the to-be-monitored grassland is determined according to the comparison result, the method comprises the following steps: When there is a historical multi-source image feature vector identical to the multi-source image feature vector in the historical image set, a historical degradation evaluation value corresponding to the historical multi-source image feature vector is taken as the degradation evaluation value; When there is no historical multi-source image feature vector identical to the multi-source image feature vector in the historical image set, a degradation evaluation base value of the to-be-monitored grassland is determined according to the grassland bareness characteristic value; The degradation evaluation base value is compensated according to the vegetation condition characteristic value and the soil erosion characteristic value to obtain the degradation evaluation value. 4.The AI image recognition-based grassland degradation level diagnosis and restoration system of claim 3, wherein, When the degradation evaluation base value of the to-be-monitored grassland is determined according to the grassland bareness characteristic value, the method comprises the following steps: The grassland bareness characteristic value is compared with a first grassland bareness characteristic value and a second grassland bareness characteristic value, and the degradation evaluation base value is determined according to the comparison result; wherein the first grassland bareness characteristic value is smaller than the second grassland bareness characteristic value; When the grassland bareness characteristic value is smaller than the first grassland bareness characteristic value, the degradation evaluation base value is determined to be a first degradation evaluation base value; When the grassland bareness characteristic value is greater than or equal to the first grassland bareness characteristic value and smaller than the second grassland bareness characteristic value, the degradation evaluation base value is determined to be a second degradation evaluation base value; When the grassland bareness characteristic value is greater than or equal to the second grassland bareness characteristic value, the degradation evaluation base value is determined to be a third degradation evaluation base value. 5.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 4, characterized in that, When the degradation evaluation base value is compensated according to the vegetation condition characteristic value and the soil erosion characteristic value to obtain the degradation evaluation value, the method comprises the following steps: The vegetation condition feature value is compared with a vegetation condition feature threshold value, the soil erosion feature value is compared with a soil erosion feature threshold value, and a compensation coefficient of the degradation evaluation base value is determined according to a comparison result; When the vegetation condition feature value is greater than the vegetation condition feature threshold value and the soil erosion feature value is greater than the soil erosion feature threshold value, the compensation coefficient is determined as a first compensation coefficient; When the vegetation condition feature value is greater than the vegetation condition feature threshold value and the soil erosion feature value is less than or equal to the soil erosion feature threshold value, the compensation coefficient is determined as a second compensation coefficient; When the vegetation condition feature value is less than or equal to the vegetation condition feature threshold value and the soil erosion feature value is greater than the soil erosion feature threshold value, the compensation coefficient is determined as a third compensation coefficient; When the vegetation condition feature value is less than or equal to the vegetation condition feature threshold value and the soil erosion feature value is less than or equal to the soil erosion feature threshold value, the compensation coefficient is determined as a fourth compensation coefficient. 6.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 5, characterized in that, When it is determined whether to adjust the degradation evaluation value according to the preset ecological indicators, the method comprises the following steps: Real-time ecological indicators of the grassland to be monitored are collected; A deviation degree of each real-time ecological indicator from a corresponding preset ecological indicator is calculated; All the deviation degrees are normalized and weighted to obtain a comprehensive ecological indicator deviation index; It is determined whether to adjust the degradation evaluation value according to the comprehensive ecological indicator deviation index; If the comprehensive ecological indicator deviation index is greater than 0.45, it is determined to adjust the degradation evaluation value; If the comprehensive ecological indicator deviation index is less than or equal to 0.45, it is determined not to adjust the degradation evaluation value. 7.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 6, characterized in that, When the adjustment coefficient of the degradation evaluation value is determined based on the root density and a final degradation evaluation value is obtained, the method comprises the following steps: A fusion vector group is constructed according to the root density and the comprehensive ecological indicator deviation index; The Euclidean distance between the fusion vector group and each historical fusion vector group is calculated, and the minimum Euclidean distance is extracted; The adjustment coefficient is determined according to the minimum Euclidean distance. 8.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 7, characterized in that, When the adjustment coefficient is determined according to the minimum Euclidean distance, the method comprises the following steps: When the minimum Euclidean distance is unique, a standard Euclidean distance is obtained, and a difference between the standard Euclidean distance and the minimum Euclidean distance is calculated, which is denoted as an Euclidean distance difference value; The Euclidean distance difference value is compared with a preset adjustment coefficient mapping table, and the adjustment coefficient is determined according to a comparison result; When the minimum Euclidean distance is not unique, an average value of all the minimum Euclidean distances is obtained, which is denoted as an average Euclidean distance, and the adjustment coefficient is determined according to the average Euclidean distance; The product of the adjustment coefficient and the degradation evaluation value is taken as the final degradation evaluation value. 9.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 8, characterized in that, When the degradation grade of the grassland to be monitored is determined according to the final degradation evaluation value, the method comprises the following steps: The final degradation evaluation value is compared with a first final degradation evaluation value and a second final degradation evaluation value, and the degradation grade is determined according to a comparison result; wherein the first final degradation evaluation value is less than the second final degradation evaluation value. determining that the degradation level is slight when the degradation evaluation end value is less than or equal to the first degradation evaluation end value, and marking the to-be-monitored grassland as a slight degradation area; determining that the degradation level is moderate when the degradation evaluation end value is greater than the first degradation evaluation end value and less than or equal to the second degradation evaluation end value, and marking the to-be-monitored grassland as a moderate degradation area; determining that the degradation level is severe when the degradation evaluation end value is greater than the second degradation evaluation end value, and marking the to-be-monitored grassland as a severe degradation area. 10.The AI image recognition-based grassland degradation level diagnosis and restoration system according to claim 9, wherein, When determining the repair strategy of the to-be-monitored grassland according to the degradation level, the method comprises: for the slight degradation area, implementing biological bud promotion technology, including applying plant growth promoters and slow-release compound fertilizers; for the moderate degradation area, implementing root cutting and bud promotion technology, using machines to cut the grass mat, and simultaneously performing water and fertilizer management; for the severe degradation area, implementing bud island planting technology.

Images