Mine ecological restoration effect evaluation system based on hyperspectrum

Abstract

The invention discloses a mine ecological restoration effect evaluation system based on hyperspectrum, and relates to the technical field of ecological environment intelligent monitoring, the system comprises a data acquisition module, a data processing module, a mine ecological restoration perception model establishment module, a vegetation restoration monitoring module, a geological restoration monitoring module and a pollution residue monitoring module; the system divides a monitoring area based on a GIS, collects hyperspectral, geological and pollution data, constructs a convolutional neural network perception model after wavelet denoising, image correction, band interpolation and spatial interpolation, and fuses multi-dimensional features to realize intelligent evaluation. The system respectively calculates an ecological restoration coefficient STX, a geological stability coefficient DZW and a pollution residual risk coefficient WRX, compares the ecological restoration coefficient STX, the geological stability coefficient DZW and the pollution residual risk coefficient WRX with corresponding threshold values, and evaluates whether the restoration effect is qualified or not; if not, a targeted repair strategy and a risk assessment report are generated; and dynamic monitoring and intelligent early warning of the ecological state of the mine are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological environment intelligent monitoring, and specifically to a mine ecological restoration effectiveness evaluation system based on hyperspectral technology. Background Art

[0002] Currently, assessments of mine ecological restoration effectiveness often rely on traditional remote sensing image comparisons, vegetation index trend analysis, or estimates based on field monitoring data from limited sample points. However, these methods have the following limitations: First, the data dimension is single, making it difficult to fully reflect overall changes in the mining area's ecosystem structure, function, and residual pollution. Second, spatial resolution and spectral information are insufficient. Traditional images and indices are prone to errors, particularly in mining environments with complex and heterogeneous coverage types, affecting assessment accuracy. Third, due to the lack of a unified indicator system and model support, assessment results often rely on expert experience or manual judgment, which is highly subjective and inconsistent, making it difficult to implement large-scale, quantitative analysis.

[0003] In addition, although some studies have attempted to introduce hyperspectral remote sensing methods, their application in actual ecological restoration assessments is still restricted due to complex processing procedures, insufficient data interpretation algorithms, or the inability to achieve fine distinction between land object categories. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a mine ecological restoration effectiveness evaluation system based on hyperspectral to solve the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a mine ecological restoration effectiveness evaluation system based on hyperspectral, including a data acquisition module, a data processing module, a mine ecological restoration perception model establishment module, a vegetation restoration monitoring module, a geological restoration monitoring module and a pollution residue monitoring module;

[0006] The data acquisition module is used to perform spatial analysis of the mine ecological restoration area using the GIS geographic information system and divide it into several monitoring areas; it collects hyperspectral data, geological data, and soil and water pollution sampling data of the monitoring area in real time;

[0007] The data processing module is used to correct image distortion through wavelet transform denoising and geometric correction algorithm, fill in missing bands using Lagrange interpolation, and complete spatial data interpolation using inverse distance weighted method;

[0008] The mine ecological restoration perception model establishment module is used to integrate hyperspectral, geological and pollution data training models based on convolutional neural networks, and optimize training through multi-dimensional feature information to achieve intelligent evaluation and output of restoration effects;

[0009] The vegetation restoration monitoring module is used to monitor the vegetation restoration in the monitoring area in real time, calculate the ecological restoration coefficient STX, and compare and analyze it with the first threshold Q1 to determine whether the ecological restoration effect of the monitoring area is qualified. If it is unqualified, a strategy is given;

[0010] The geological restoration monitoring module is used to monitor the geological structure characteristics of the monitoring area in real time, calculate the geological stability coefficient DZW, and compare and analyze it with the second threshold Q2 to determine whether the stability of the geological structure repaired in the monitoring area is qualified. If it is unqualified, a strategy is given;

[0011] The pollution residue monitoring module is used to monitor the concentration of heavy metal elements in the soil and water indicators in real time, calculate the pollution residue risk coefficient WRX, and compare and analyze it with the third threshold Q3 to determine whether the water and soil pollution remediation indicators in the monitored area are qualified. If not, a strategy is given.

[0012] Preferably, the data acquisition module includes a division unit, a hyperspectral acquisition unit, a geological monitoring acquisition unit and a soil and water pollution sampling unit;

[0013] The division unit is used to perform spatial analysis using a GIS geographic information system and define the boundaries of the mine ecological restoration area, divide the mine ecological restoration area into several monitoring areas, and locate and mark each monitoring area on an electronic distribution map, marked as: Bp1, Bp2, ..., BpZ; Z represents the number of monitoring areas;

[0014] The hyperspectral acquisition unit is used to use a hyperspectral imager mounted on a drone to collect raw spectral reflectance data of multiple bands from 400 to 2500 nm in the visible to short-wave infrared range of the monitoring area, including red band, near-infrared band, short-wave infrared band and red-edge band;

[0015] The geological monitoring acquisition unit is used to use a ground laser radar LiDAR and a three-dimensional photogrammeter to collect the slope angle, slope length and crack width of the monitoring area; and use a portable geotechnical in-situ shear meter to collect the soil shear strength of the monitoring area;

[0016] The soil and water pollution sampling unit is used to collect the concentration data of heavy metal elements in the soil of the monitoring area using a portable XRF heavy metal detector; and to collect the index data of pH value, ammonia nitrogen and chemical oxygen demand of the water body in the monitoring area using an online water quality monitor.

[0017] Preferably, the data processing module is used to use wavelet transform technology to perform band denoising on the original hyperspectral band data; use a geometric correction algorithm to perform geometric distortion correction on the hyperspectral image to achieve spatial alignment; use Lagrange interpolation technology to interpolate the missing band data in the hyperspectral image; and use an inverse distance weighted method to perform spatial interpolation and completion on the spatial monitoring point data.

[0018] Preferably, the mine ecological restoration perception model establishment module is used to use a convolutional neural network to construct the initial network structure of the mine ecological restoration perception model, and train and test the initial model with hyperspectral data, geological monitoring data and soil and water pollution sampling data; the trained convolutional neural network model is used as the mine ecological restoration perception model; at the same time, the multidimensional feature information composed of the vegetation coverage index feature vector, the geological stability index feature vector and the pollution risk index feature vector is used as the intermediate layer input to extract its feature information, and the feature information is used again to train and test the mine ecological restoration perception model. The trained mine ecological restoration perception model outputs an assessment of mine ecological restoration in real time.

[0019] Preferably, the vegetation restoration monitoring module includes a first calculation unit and a first analysis unit;

[0020] The first calculation unit is used to monitor the vegetation restoration in the monitoring area in real time, and calculates the ecological restoration coefficient STX after dimensionless processing based on the reflectance data of the hyperspectral remote sensing image. The formula is as follows:

[0021] STX=w1*FVC+w2*CIre+w3*NDMI+w4*SHI;

[0022] Where, FVC represents vegetation cover, CIre represents red-edge chlorophyll index, NDMI represents vegetation moisture index, SHI represents species diversity index, w1, w2, w3 and w4 represent weight coefficients;

[0023]

[0024] Where NDVI stands for Normalized Difference Vegetation Index, s Normalized Difference Vegetation Index (NDVI) for pure bare land v represents the normalized difference vegetation index of pure vegetation, ρNIR represents the near-infrared band reflectance, and ρRED represents the red band reflectance;

[0025]

[0026] Where, ρre represents the red edge band reflectivity;

[0027]

[0028] Where, ρSWIR represents the reflectivity in the short-wave infrared band;

[0029]

[0030] Where n represents the number of species identified by hyperspectral analysis, and p i represents the proportion of species in category i identified by hyperspectral analysis.

[0031] Preferably, the first analysis unit is used to preset a first threshold Q1 in advance, and compare and analyze the ecological restoration coefficient STX with the first threshold Q1, and obtaining the first evaluation result includes:

[0032] When the ecological restoration coefficient STX ≥ the first threshold Q1, it indicates that the ecological restoration effect of the monitored area is qualified, there is no risk of surface instability due to poor vegetation growth and a single ecological structure, and continuous monitoring is required;

[0033] When the ecological restoration coefficient STX is less than the first threshold Q1, it means that the ecological restoration effect in the monitored area is unsatisfactory, and there is a risk of surface instability due to poor vegetation growth and a single ecological structure. The first early warning instruction is triggered and the first strategy is generated: increase plant diversity by 30%, increase coverage density by 20%, introduce highly adaptable pioneer plants and ground cover plants; and generate a vegetation risk assessment report.

[0034] Preferably, the geological remediation monitoring module includes a second calculation unit and a second analysis unit;

[0035] The second calculation unit is used to monitor the geological structure characteristics of the monitoring area in real time, and calculate the geological stability coefficient DZW after dimensionless processing based on the slope angle, slope length, crack width and soil shear strength. The formula is as follows:

[0036]

[0037] Where θ represents the slope gradient, Rs represents the soil shear strength, Rc represents the shear strength of the repair and reinforcement material, Dc represents the crack width, Ls represents the slope length, and a1, a2, and a3 represent weight coefficients.

[0038] Preferably, the second analysis unit is used to preset a second threshold Q2 in advance, and compare and analyze the geological stability coefficient DZW with the second threshold Q2, and obtain the second evaluation result including:

[0039] When the geological stability coefficient DZW ≥ the second threshold Q2, it means that the stability of the repaired geological structure in the monitoring area is qualified, there is no risk of geological landslide and collapse, and continuous monitoring is required;

[0040] When the geological stability coefficient DZW is less than the second threshold Q2, it means that the stability of the repaired geological structure in the monitored area is unqualified and there is a risk of geological landslide and collapse. The second early warning instruction is triggered and the second strategy is generated: for structurally unstable geology, 1.5 times the original design support piles or anchor piles are added, and the buried depth of the support piles or anchor piles is increased to 130% of the current depth; a geological risk assessment report is generated.

[0041] Preferably, the pollution residue monitoring module includes a third calculation unit and a third analysis unit;

[0042] The third calculation unit is used to monitor the concentration data of heavy metal elements in the soil and the index data of pH value, ammonia nitrogen and chemical oxygen demand of the water body in real time, and extract the purified soil spectrum residual value from the hyperspectral remote sensing image. After dimensionless processing, the pollution residual risk coefficient WRX is calculated. The formula is as follows:

[0043]

[0044] In the formula, m represents the type of heavy metals in the soil, C j Indicates the concentration of the jth heavy metal in the soil, Cth j represents the environmental protection threshold of the jth heavy metal, Rnd represents the residual value of the purified soil spectrum, Rref represents the standard reference spectrum value, WPI represents the water pollution index, d1, d2 and d3 represent the weight coefficients respectively;

[0045]

[0046] In the formula, g represents the number of hyperspectral bands involved in the calculation, Robs u Indicates the reflectivity in the uth band, Rref u represents the standard reference reflectance in the u-th band;

[0047] WPI=γ1*f(xph)+γ2*f(xcod)+γ3*f(xnh);

[0048] Where xph represents the pH value of the water body, xcod represents the concentration of organic matter in the water body, xnh represents the concentration of ammonia nitrogen in the water body, and γ1, γ2 and γ3 represent weight coefficients.

[0049] Preferably, the third analysis unit is used to preset a third threshold value Q3 in advance, and compare and analyze the pollution residual risk coefficient WRX with the third threshold value Q3, and obtain the third evaluation result including:

[0050] When the residual pollution risk coefficient WRX is less than the third threshold Q3, it means that the water and soil pollution remediation indicators in the monitoring area are qualified, there is no risk of residual pollution, and monitoring is continued;

[0051] When the pollution residual risk coefficient WRX ≥ the third threshold Q3, it means that the water and soil pollution remediation indicators in the monitored area are unqualified and there is a risk of pollution residue, which triggers the third early warning instruction and generates the third strategy: introduce plant remediation technology, plant lead and zinc-enriched plants, carry out synergistic adsorption, supplemented by soil chemical stabilizers including bentonite and ferrous sulfate to inhibit the activity of heavy metals; deploy microbial purification materials and ecological floating beds in areas with abnormal COD and ammonia nitrogen indicators in the water body; and generate a pollution residual risk assessment report.

[0052] The present invention provides a mine ecological restoration effectiveness evaluation system based on hyperspectral data. It has the following beneficial effects:

[0053] (1) This mine ecological restoration effectiveness evaluation system based on hyperspectral technology uses a drone equipped with a hyperspectral imager to conduct multi-band remote sensing monitoring of the mining area in the range of 400–2500 nm, achieving high-precision, non-destructive vegetation and soil cover detection in large-scale areas. At the same time, it combines LiDAR laser radar with multi-source online acquisition equipment to quickly obtain geological and pollution information, greatly improving monitoring efficiency. It is suitable for ecological restoration work in complex environments such as mountainous areas and uninhabited areas.

[0054] (2) This hyperspectral-based mine ecological restoration effectiveness evaluation system, by integrating hyperspectral data, geological structure data and soil and water pollution data, combines convolutional neural networks to establish a mine ecological restoration perception model, and realizes deep learning and joint modeling of multi-dimensional ecological indicators in the restoration area, greatly improving the accuracy and robustness of the evaluation model. It can intelligently judge the effectiveness of ecological restoration and automatically generate targeted strategies, overcoming the limitations of traditional reliance on a single indicator or manual experience judgment.

[0055] (3) The mine ecological restoration effectiveness evaluation system based on hyperspectral data constructs three core indicators: vegetation restoration coefficient STX, geological stability coefficient DZW and pollution residual risk coefficient WRX. By comparing with the preset thresholds Q1, Q2 and Q3, it realizes intelligent monitoring, risk warning and result traceability of key dimensions such as ecological restoration, geological safety and pollution control, provides a quantitative basis and traceability mechanism for ecological environment supervision, and significantly improves the monitoring efficiency and management scientificity.

[0056] (4) The mine ecological restoration effectiveness evaluation system based on hyperspectral can automatically generate targeted ecological intervention strategies for different types of unqualified evaluation results, including: poor ecological recovery, geological instability, and high pollution residues. These strategies include increasing vegetation diversity, strengthening structures, and introducing restoration plants, as well as risk assessment reports. This provides decision-making basis and action recommendations for subsequent governance, effectively promoting the closed-loop execution and dynamic optimization of restoration work. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1This is a block diagram and flow chart of the mine ecological restoration effectiveness evaluation system based on hyperspectral technology of the present invention. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] Example 1

[0060] See also Figure 1 The present invention provides a mine ecological restoration effectiveness evaluation system based on hyperspectral, including a data acquisition module, a data processing module, a mine ecological restoration perception model establishment module, a vegetation restoration monitoring module, a geological restoration monitoring module and a pollution residue monitoring module;

[0061] The data acquisition module is used to perform spatial analysis of the mine ecological restoration area using the GIS geographic information system and divide it into several monitoring areas; it collects hyperspectral data, geological data, and soil and water pollution sampling data of the monitoring area in real time;

[0062] The data processing module is used to correct image distortion through wavelet transform denoising and geometric correction algorithm, fill in missing bands using Lagrange interpolation, and complete spatial data interpolation using inverse distance weighted method;

[0063] The mine ecological restoration perception model establishment module is used to integrate hyperspectral, geological and pollution data training models based on convolutional neural networks, and optimize training through multi-dimensional feature information to achieve intelligent evaluation and output of restoration effects;

[0064] The vegetation restoration monitoring module is used to monitor the vegetation restoration in the monitoring area in real time, calculate the ecological restoration coefficient STX, and compare and analyze it with the first threshold Q1 to determine whether the ecological restoration effect of the monitoring area is qualified. If it is unqualified, a strategy is given;

[0065] The geological restoration monitoring module is used to monitor the geological structure characteristics of the monitoring area in real time, calculate the geological stability coefficient DZW, and compare and analyze it with the second threshold Q2 to determine whether the stability of the geological structure repaired in the monitoring area is qualified. If it is unqualified, a strategy is given;

[0066] The pollution residue monitoring module is used to monitor the concentration of heavy metal elements in the soil and water indicators in real time, calculate the pollution residue risk coefficient WRX, and compare and analyze it with the third threshold Q3 to determine whether the water and soil pollution remediation indicators in the monitored area are qualified. If not, a strategy is given.

[0067] In this embodiment, by integrating hyperspectral imaging, GIS geographic information system and multi-source monitoring data, a multidimensional ecological perception model including vegetation restoration, geological structure and pollution residues was constructed. It not only achieved a comprehensive evaluation of the ecological restoration effect of the mine, but also dynamically monitored the restoration results in different dimensions and provided strategic feedback by setting multiple threshold judgment criteria, thereby significantly improving the scientificity, real-timeness and intelligence of the mine ecological restoration project.

[0068] Example 2

[0069] This embodiment is explained in Example 1. Specifically, the data acquisition module includes a division unit, a hyperspectral acquisition unit, a geological monitoring acquisition unit, and a soil and water pollution sampling unit;

[0070] The division unit is used to perform spatial analysis using a GIS geographic information system and define the boundaries of the mine ecological restoration area, divide the mine ecological restoration area into several monitoring areas, and locate and mark each monitoring area on an electronic distribution map, marked as: Bp1, Bp2, ..., BpZ; Z represents the number of monitoring areas;

[0071] The hyperspectral acquisition unit is used to use a hyperspectral imager mounted on a drone to collect raw spectral reflectance data of multiple bands from 400 to 2500 nm in the visible to short-wave infrared range of the monitoring area, including red band, near-infrared band, short-wave infrared band and red-edge band;

[0072] The geological monitoring acquisition unit is used to use a ground laser radar LiDAR and a three-dimensional photogrammeter to collect the slope angle, slope length and crack width of the monitoring area; and use a portable geotechnical in-situ shear meter to collect the soil shear strength of the monitoring area;

[0073] The soil and water pollution sampling unit is used to collect the concentration data of heavy metal elements in the soil of the monitoring area using a portable XRF heavy metal detector; and to collect the index data of pH value, ammonia nitrogen and chemical oxygen demand of the water body in the monitoring area using an online water quality monitor.

[0074] In this embodiment, by setting up division units and multi-type collection units, high-precision spatial division of the mine ecological restoration area and accurate acquisition of multi-dimensional environmental information are achieved. It can comprehensively collect multi-source original data including spectral reflectance, geological structure parameters and soil and water pollution indicators, providing detailed and reliable basic data support for subsequent ecological perception modeling and restoration effectiveness evaluation, greatly improving the comprehensiveness and field adaptability of data acquisition.

[0075] Example 3

[0076] This embodiment is explained in Example 2. Specifically, the data processing module is used to use wavelet transform technology to perform band denoising on the original hyperspectral band data; use a geometric correction algorithm to perform geometric distortion correction on the hyperspectral image to achieve spatial alignment; use Lagrange interpolation technology to interpolate the missing band data in the hyperspectral image; and use the inverse distance weighted method to perform spatial interpolation and completion on the spatial monitoring point data.

[0077] In this embodiment, by adopting advanced data processing technologies such as wavelet transform technology, geometric correction algorithm, Lagrange interpolation and inverse distance weighted method, the quality and accuracy of hyperspectral image data are effectively improved, noise and distortion are eliminated, missing bands are supplemented, and the integrity and accuracy of spatial monitoring data are ensured, thereby providing more accurate basic data for the evaluation of mine ecological restoration effects.

[0078] Example 4

[0079] This embodiment is an explanation of Embodiment 3. Specifically, the mine ecological restoration perception model establishment module is used to use a convolutional neural network to construct the initial network structure of the mine ecological restoration perception model, and train and test the initial model with hyperspectral data, geological monitoring data and soil and water pollution sampling data; the trained convolutional neural network model is used as the mine ecological restoration perception model; at the same time, the multidimensional feature information composed of the vegetation coverage index feature vector, the geological stability index feature vector and the pollution risk index feature vector is used as the intermediate layer input to extract its feature information, and the feature information is used again to train and test the mine ecological restoration perception model. The trained mine ecological restoration perception model outputs an assessment of mine ecological restoration in real time.

[0080] In this example, a convolutional neural network-based mine ecological restoration perception model was constructed. This model, combined with hyperspectral data, geological monitoring data, and soil and water pollution sampling data, was used to extract and train multidimensional feature information. This approach effectively enhanced the model's intelligence and enabled real-time output of mine ecological restoration assessment results.

[0081] Example 5

[0082] This embodiment is explained in Example 4. Specifically, the vegetation restoration monitoring module includes a first calculation unit and a first analysis unit;

[0083] The first calculation unit is used to monitor the vegetation restoration in the monitoring area in real time, and calculates the ecological restoration coefficient STX after dimensionless processing based on the reflectance data of the hyperspectral remote sensing image. The formula is as follows:

[0084] STX=w1*FVC+w2*CIre+w3*NDMI+w4*SHI;

[0085] Where FVC represents vegetation coverage, CIre represents red-edge chlorophyll index, NDMI represents vegetation moisture index, SHI represents species diversity index, w1, w2, w3 and w4 represent weight coefficients, 0<w1<1, 0<w2<1, 0<w3<1, 0<w4<1 and w1+w2+w3+w4=1;

[0086]

[0087] Where NDVI stands for Normalized Difference Vegetation Index, s Normalized Difference Vegetation Index (NDVI) for pure bare land v represents the normalized difference vegetation index of pure vegetation, ρNIR represents the near-infrared band reflectance, and ρRED represents the red band reflectance;

[0088]

[0089] Where, ρre represents the red edge band reflectivity;

[0090]

[0091] Where, ρSWIR represents the reflectivity in the short-wave infrared band;

[0092]

[0093] Where n represents the number of species identified by hyperspectral analysis, and p i represents the proportion of species in category i identified by hyperspectral analysis.

[0094] In this example, the vegetation restoration monitoring module, combined with reflectance data from hyperspectral remote sensing images, utilizes dimensionless processing and multiple exponential calculations to calculate and obtain the ecological restoration coefficient (STX) in real time, effectively enabling accurate monitoring and assessment of vegetation restoration progress. This method comprehensively considers multidimensional factors such as vegetation coverage, species diversity, and reflectance across various bands, providing a more comprehensive and accurate assessment of ecological restoration status. This provides a scientific basis for ecological restoration decision-making and strategy adjustments, significantly improving the management efficiency of mine ecological restoration.

[0095] Example 6

[0096] This embodiment is an explanation of Embodiment 5. Specifically, the first analysis unit is used to preset a first threshold value Q1 in advance, and compare and analyze the ecological restoration coefficient STX with the first threshold value Q1. Obtaining a first evaluation result includes:

[0097] When the ecological restoration coefficient STX ≥ the first threshold Q1, it indicates that the ecological restoration effect of the monitored area is qualified, there is no risk of surface instability due to poor vegetation growth and a single ecological structure, and continuous monitoring is required;

[0098] When the ecological restoration coefficient STX is less than the first threshold Q1, it means that the ecological restoration effect in the monitored area is unsatisfactory, and there is a risk of surface instability due to poor vegetation growth and a single ecological structure. The first early warning instruction is triggered and the first strategy is generated: increase plant diversity by 30%, increase coverage density by 20%, introduce highly adaptable pioneer plants and ground cover plants; and generate a vegetation risk assessment report.

[0099] In this embodiment, the first analysis unit of the vegetation restoration monitoring module compares the ecological restoration coefficient STX with a preset first threshold Q1, accurately assessing the ecological restoration effect in the monitored area. When the ecological restoration coefficient falls below the threshold, the system triggers a timely warning and generates a corresponding strategy, proposing specific measures such as increasing plant diversity and improving cover density to ensure continued improvement in vegetation restoration. This intelligent warning and strategy generation mechanism not only effectively prevents the risk of surface instability but also provides scientific and actionable guidance for ecological restoration.

[0100] Example 7

[0101] This embodiment is explained in Example 4. Specifically, the geological restoration monitoring module includes a second calculation unit and a second analysis unit;

[0102] The second calculation unit is used to monitor the geological structure characteristics of the monitoring area in real time, and calculate the geological stability coefficient DZW after dimensionless processing based on the slope angle, slope length, crack width and soil shear strength. The formula is as follows:

[0103]

[0104] Where θ represents the slope gradient, Rs represents the soil shear strength, Rc represents the shear strength of the repair and reinforcement material, Dc represents the crack width, Ls represents the slope length, a1, a2, and a3 represent weight coefficients, 0<a1<1, 0<a2<1, 0<a3<1, and a1+a2+a3=1.

[0105] In this embodiment, the second calculation unit of the geological restoration monitoring module combines key geological characteristics such as slope angle, crack width, and soil shear strength to calculate the geological stability coefficient DZW in real time, enabling a comprehensive assessment of the geological restoration effect in the monitored area. This technology effectively measures the stability of geological structures during the restoration process, providing a basis for timely identification of potential geological risks, ensuring geological safety during mine ecological restoration, preventing geological disasters, and helping to optimize restoration strategies.

[0106] Example 8

[0107] This embodiment is explained in Example 7. Specifically, the second analysis unit is used to preset a second threshold value Q2 in advance, and compare and analyze the geological stability coefficient DZW with the second threshold value Q2. Obtaining the second evaluation result includes:

[0108] When the geological stability coefficient DZW ≥ the second threshold Q2, it means that the stability of the repaired geological structure in the monitoring area is qualified, there is no risk of geological landslide and collapse, and continuous monitoring is required;

[0109] When the geological stability coefficient DZW is less than the second threshold Q2, it means that the stability of the repaired geological structure in the monitored area is unqualified and there is a risk of geological landslide and collapse. The second early warning instruction is triggered and the second strategy is generated: for structurally unstable geology, 1.5 times the original design support piles or anchor piles are added, and the buried depth of the support piles or anchor piles is increased to 130% of the current depth; a geological risk assessment report is generated.

[0110] In this embodiment, the second analysis unit of the geological restoration monitoring module compares the geological stability coefficient DZW with a preset second threshold Q2, enabling real-time assessment of the geological stability of the monitored area. When the geological stability coefficient is detected to be below the preset threshold, the system quickly triggers an early warning and provides targeted restoration strategies, such as adding support piles or anchor piles, to prevent the risk of landslides or collapses. This technology not only improves the safety of mine ecological restoration but also effectively prevents ecological restoration failures caused by geological instability.

[0111] Example 9

[0112] This embodiment is explained in Example 4. Specifically, the pollution residue monitoring module includes a third calculation unit and a third analysis unit;

[0113] The third calculation unit is used to monitor the concentration data of heavy metal elements in the soil and the index data of pH value, ammonia nitrogen and chemical oxygen demand of the water body in real time, and extract the purified soil spectrum residual value from the hyperspectral remote sensing image. After dimensionless processing, the pollution residual risk coefficient WRX is calculated. The formula is as follows:

[0114]

[0115] In the formula, m represents the type of heavy metals in the soil, C j Indicates the concentration of the jth heavy metal in the soil, Cth j represents the environmental threshold of the jth heavy metal, Rnd represents the residual value of the purified soil spectrum, Rref represents the standard reference spectrum value, WPI represents the water pollution index, d1, d2 and d3 represent weight coefficients, 0<d1<1, 0<d2<1, 0<d3<1, and d1+d2+d3=1;

[0116]

[0117] In the formula, g represents the number of hyperspectral bands involved in the calculation, Robs u Indicates the reflectivity in the uth band, Rref u represents the standard reference reflectance in the u-th band;

[0118] WPI=γ1*f(xph)+γ2*f(xcod)+γ3*f(xnh);

[0119] Where xph represents the pH value of the water body, xcod represents the concentration of organic matter in the water body, xnh represents the concentration of ammonia nitrogen in the water body, γ1, γ2 and γ3 represent weight coefficients, 0<γ1<1, 0<γ2<1, 0<γ3<1, and γ1+γ2+γ3=1.

[0120] In this embodiment, the third analysis unit of the residual pollution monitoring module combines soil heavy metal concentrations, soil spectral residual values, and water pollution indicators to calculate the residual pollution risk coefficient WRX in real time, enabling a comprehensive assessment of the pollution risk in the mine ecological restoration area. When the residual pollution risk coefficient exceeds a preset threshold, the system can quickly issue an alarm and formulate targeted remediation strategies, such as strengthening soil heavy metal control and improving water pollution treatment. This technology effectively prevents the impact of pollutants on the ecological restoration process.

[0121] Example 10

[0122] This embodiment is explained in Example 9. Specifically, the third analysis unit is used to preset a third threshold value Q3 in advance, and compare and analyze the pollution residual risk coefficient WRX with the third threshold value Q3. Obtaining a third evaluation result includes:

[0123] When the residual pollution risk coefficient WRX is less than the third threshold Q3, it means that the water and soil pollution remediation indicators in the monitoring area are qualified, there is no risk of residual pollution, and monitoring is continued;

[0124] When the pollution residual risk coefficient WRX ≥ the third threshold Q3, it means that the water and soil pollution remediation indicators in the monitored area are unqualified and there is a risk of pollution residue, which triggers the third early warning instruction and generates the third strategy: introduce plant remediation technology, plant lead and zinc-enriched plants, carry out synergistic adsorption, supplemented by soil chemical stabilizers including bentonite and ferrous sulfate to inhibit the activity of heavy metals; deploy microbial purification materials and ecological floating beds in areas with abnormal COD and ammonia nitrogen indicators in the water body; and generate a pollution residual risk assessment report.

[0125] In this embodiment, the third analysis unit of the residual pollution monitoring module, combined with a comparative analysis of the residual pollution risk coefficient WRX and a preset threshold Q3, enables real-time assessment of the residual pollution risk in the remediation area. When the pollution risk exceeds the threshold, the system automatically triggers an early warning and formulates an effective remediation strategy, such as introducing phytoremediation technology and microbial purification materials to ensure effective control of soil and water pollution. This technology provides accurate risk assessment and timely intervention measures for mine ecological restoration, significantly improving the scientific nature of remediation efforts and the sustainability of environmental recovery.

[0126] The threshold value is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each set of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0127] The above formulas are obtained by collecting a large amount of data and performing software simulation, and a formula close to the actual value is selected. The coefficients in the formula are set by those skilled in the art according to actual conditions. The above is only a preferred specific implementation method of the present invention, but the protection scope of the present invention is not limited to this. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, can make equivalent replacements or changes based on the technical solution and inventive concept of the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The mine ecological restoration effectiveness evaluation system based on hyperspectral is characterized by: It includes data acquisition module, data processing module, mine ecological restoration perception model establishment module, vegetation restoration monitoring module, geological restoration monitoring module and pollution residue monitoring module; The data acquisition module is used to perform spatial analysis of the mine ecological restoration area using the GIS geographic information system and divide it into several monitoring areas; it collects hyperspectral data, geological data, and soil and water pollution sampling data of the monitoring area in real time; The data processing module is used to correct image distortion through wavelet transform denoising and geometric correction algorithm, fill in missing bands using Lagrange interpolation, and complete spatial data interpolation using inverse distance weighted method; The mine ecological restoration perception model establishment module is used to integrate hyperspectral, geological and pollution data training models based on convolutional neural networks, and optimize training through multi-dimensional feature information to achieve intelligent evaluation and output of restoration effects; The vegetation restoration monitoring module is used to monitor the vegetation restoration in the monitoring area in real time, calculate the ecological restoration coefficient STX, and compare and analyze it with the first threshold Q1 to determine whether the ecological restoration effect of the monitoring area is qualified. If it is unqualified, a strategy is given; The geological restoration monitoring module is used to monitor the geological structure characteristics of the monitoring area in real time, calculate the geological stability coefficient DZW, and compare and analyze it with the second threshold Q2 to determine whether the stability of the geological structure repaired in the monitoring area is qualified. If it is unqualified, a strategy is given; The pollution residue monitoring module is used to monitor the concentration of heavy metal elements in the soil and water indicators in real time, calculate the pollution residue risk coefficient WRX, and compare and analyze it with the third threshold Q3 to determine whether the water and soil pollution remediation indicators in the monitored area are qualified. If not, a strategy is given.

2. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 1 is characterized in that: The data acquisition module includes a division unit, a hyperspectral acquisition unit, a geological monitoring acquisition unit and a soil and water pollution sampling unit; The division unit is used to perform spatial analysis using a GIS geographic information system and define the boundaries of the mine ecological restoration area, divide the mine ecological restoration area into several monitoring areas, and locate and mark each monitoring area on an electronic distribution map, marked as: Bp1, Bp2, ..., BpZ; Z represents the number of monitoring areas; The hyperspectral acquisition unit is used to use a hyperspectral imager mounted on a drone to collect raw spectral reflectance data of multiple bands from 400 to 2500 nm in the visible to short-wave infrared range of the monitoring area, including red band, near-infrared band, short-wave infrared band and red-edge band; The geological monitoring acquisition unit is used to use a ground laser radar LiDAR and a three-dimensional photogrammeter to collect the slope angle, slope length and crack width of the monitoring area; and use a portable geotechnical in-situ shear meter to collect the soil shear strength of the monitoring area; The soil and water pollution sampling unit is used to collect the concentration data of heavy metal elements in the soil of the monitoring area using a portable XRF heavy metal detector; and to collect the index data of pH value, ammonia nitrogen and chemical oxygen demand of the water body in the monitoring area using an online water quality monitor.

3. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 2 is characterized in that: The data processing module is used to use wavelet transform technology to perform band denoising on the original hyperspectral band data; use a geometric correction algorithm to correct the geometric distortion of the hyperspectral image to achieve spatial alignment; use Lagrange interpolation technology to interpolate the missing band data in the hyperspectral image; and use an inverse distance weighted method to perform spatial interpolation and completion on the spatial monitoring point data.

4. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 3 is characterized in that: The mine ecological restoration perception model establishment module is used to use a convolutional neural network to construct the initial network structure of the mine ecological restoration perception model, and train and test the initial model with hyperspectral data, geological monitoring data, and soil and water pollution sampling data; the trained convolutional neural network model is used as the mine ecological restoration perception model; at the same time, the multidimensional feature information composed of the vegetation coverage index feature vector, the geological stability index feature vector, and the pollution risk index feature vector is used as the intermediate layer input to extract its feature information. The feature information is used again to train and test the mine ecological restoration perception model. The trained mine ecological restoration perception model outputs an assessment of mine ecological restoration in real time.

5. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 4 is characterized in that: The vegetation restoration monitoring module includes a first calculation unit and a first analysis unit; The first calculation unit is used to monitor the vegetation restoration in the monitoring area in real time, and calculates the ecological restoration coefficient STX after dimensionless processing based on the reflectance data of the hyperspectral remote sensing image. The formula is as follows: Where, represents vegetation coverage, represents the red-edge chlorophyll index, represents the vegetation moisture index, represents the species diversity index, w1, w2, w3 and w4 represent weight coefficients; Where, represents the normalized difference vegetation index, represents the normalized difference vegetation index of pure bare land, represents the normalized difference vegetation index of pure vegetation, represents the reflectivity in the near-infrared band, Indicates the red band reflectivity; Where, Indicates the red edge band reflectivity; Where, Indicates the reflectivity of the shortwave infrared band; Where n represents the number of species identified by hyperspectral analysis, represents the proportion of species in category i identified by hyperspectral analysis.

6. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 5 is characterized in that: The first analysis unit is used to preset a first threshold Q1 in advance, and compare and analyze the ecological restoration coefficient STX with the first threshold Q1 to obtain a first evaluation result including: When the ecological restoration coefficient STX ≥ the first threshold Q1, it indicates that the ecological restoration effect of the monitored area is qualified, there is no risk of surface instability due to poor vegetation growth and a single ecological structure, and continuous monitoring is required; When the ecological restoration coefficient STX is less than the first threshold Q1, it means that the ecological restoration effect in the monitored area is unsatisfactory. There is a risk of surface instability due to poor vegetation growth and a single ecological structure. The first early warning instruction is triggered and the first strategy is generated: increase plant diversity by 30%, increase coverage density by 20%, introduce highly adaptable pioneer plants and ground cover plants; and generate a vegetation risk assessment report.

7. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 4 is characterized in that: The geological restoration monitoring module includes a second calculation unit and a second analysis unit; The second calculation unit is used to monitor the geological structure characteristics of the monitoring area in real time, and calculate the geological stability coefficient DZW after dimensionless processing based on the slope angle, slope length, crack width and soil shear strength. The formula is as follows: Where, represents the slope gradient, Rs represents the soil shear strength, and Rc represents the shear strength of the repair and reinforcement material. represents the crack width, represents the slope length, and a1, a2, and a3 represent weight coefficients.

8. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 7 is characterized in that: The second analysis unit is used to preset a second threshold value Q2 in advance, and compare and analyze the geological stability coefficient DZW with the second threshold value Q2 to obtain a second evaluation result including: When the geological stability coefficient DZW ≥ the second threshold Q2, it means that the stability of the repaired geological structure in the monitoring area is qualified, there is no risk of geological landslide and collapse, and continuous monitoring is required; When the geological stability coefficient DZW is less than the second threshold Q2, it means that the stability of the repaired geological structure in the monitored area is unqualified and there is a risk of geological landslide and collapse. The second early warning instruction is triggered and the second strategy is generated: for structurally unstable geology, 1.5 times the original design support piles or anchor piles are added, and the buried depth of the support piles or anchor piles is increased to 130% of the current depth; a geological risk assessment report is generated.

9. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 4 is characterized in that: The pollution residue monitoring module includes a third calculation unit and a third analysis unit; The third calculation unit is used to monitor the concentration data of heavy metal elements in the soil and the index data of pH value, ammonia nitrogen and chemical oxygen demand of the water body in real time, and extract the purified soil spectrum residual value from the hyperspectral remote sensing image. After dimensionless processing, the pollution residual risk coefficient WRX is calculated. The formula is as follows: In the formula, m represents the type of heavy metals in the soil, represents the concentration of the jth heavy metal in the soil, represents the environmental protection threshold of the jth heavy metal, represents the residual value of the purified soil spectrum, represents the standard reference spectrum value, WPI represents the water pollution index, 、 and Respectively represent weight coefficients; In the formula, g represents the number of hyperspectral bands involved in the calculation, represents the reflectivity in the u-th band, represents the standard reference reflectance in the u-th band; Where, Indicates the pH value of water. Indicates the concentration of organic matter in water. Indicates the concentration of ammonia nitrogen in water. 、 and Represents the weight coefficient.

10. The mine ecological restoration effectiveness evaluation system based on hyperspectral according to claim 9 is characterized in that: The third analysis unit is used to preset a third threshold value Q3 in advance, and compare and analyze the pollution residual risk coefficient WRX with the third threshold value Q3 to obtain a third evaluation result including: When the residual pollution risk coefficient WRX is less than the third threshold Q3, it means that the water and soil pollution remediation indicators in the monitoring area are qualified, there is no risk of residual pollution, and monitoring is continued; When the pollution residual risk coefficient WRX ≥ the third threshold Q3, it means that the water and soil pollution remediation indicators in the monitored area are unqualified and there is a risk of pollution residue, which triggers the third early warning instruction and generates the third strategy: introduce plant remediation technology, plant lead and zinc-enriched plants, carry out synergistic adsorption, supplemented by soil chemical stabilizers including bentonite and ferrous sulfate to inhibit the activity of heavy metals; deploy microbial purification materials and ecological floating beds in areas with abnormal COD and ammonia nitrogen indicators in the water body; and generate a pollution residual risk assessment report.

Images