Geographic space grid maturity evaluation optimization method based on big data

Abstract

The invention provides a geographic space grid maturity evaluation optimization method based on big data, and the method specifically comprises the steps: obtaining the priori data, such as a meteorological station or a satellite cloud picture, to position a cloud coverage area, and carrying out the cloud mask identification of a night optical image; a multi-temporal comparison method is adopted to carry out brightness repairing on a cloud shielding area, and when cloud coverage occurs for multiple continuous days, high-resolution microwave radar remote sensing data or ground sensor data can be called to carry out auxiliary compensation; according to the method, an attenuation coefficient of a cloud layer to an optical signal is calculated based on a cloud optical thickness model in a scene capable of penetrating through the cloud layer, then quantitative gain correction is carried out on pixel-level brightness of a night image, and after a cloud coverage area is effectively corrected and recovered, better geospatial grid maturity evaluation is realized.

Description

Technical Field

[0001] The present invention relates to the field of geographic information processing technology, and in particular to a geographic space grid maturity evaluation optimization method based on big data. Background Art

[0002] Existing methods for grid assessment are often affected by adverse weather conditions or thick cloud cover, resulting in insufficient or missing satellite image clarity, which in turn reduces the accuracy of regional economic activities, social development, and environmental conditions. Therefore, a cloud interference detection and compensation mechanism is needed that can integrate multi-source remote sensing imagery with meteorological prior information in a big data context to ensure the validity and accuracy of the images used for grid maturity assessment.

[0003] In nighttime remote sensing observations and urban lighting analysis, clouds often interfere with acquired images, especially when cloud cover is high or nighttime light is low. Traditional brightness correction methods struggle to accurately separate clouds from actual ground brightness, thus affecting the correct assessment of nighttime ground features, urban lighting distribution, and ambient light pollution. Existing technologies typically rely on single-phase infrared or visible light images for judgment and correction, or only perform rough corrections in conjunction with meteorological data. This fails to take into account the comprehensive utilization of multi-source data, resulting in deficiencies in cloud edge recognition, brightness estimation under continuous cloud cover, and quantitative correction of cloud-penetrable areas. Furthermore, with the continuous enrichment of observation satellite and ground-based sensor data, how to fully integrate multi-phase remote sensing images, meteorological cloud map prior information, and backup microwave data to improve the detection, correction, and brightness reconstruction of nighttime cloud interference has become a pressing technical issue. Summary of the invention

[0004] The present invention aims to at least solve the technical problems existing in the prior art, and in particular innovatively proposes a geospatial grid maturity evaluation optimization method based on big data, comprising:

[0005] S1, based on prior detection of meteorological data, obtains information from meteorological stations or satellite cloud images in real time and masks the cloud-covered area in the night light image;

[0006] S2, dynamic selection of historical reference data, optimal selection for compensating for image history effects to improve image accuracy;

[0007] S3, multi-temporal phase comparison correction, cross-comparison of multiple night light images of the same area in similar time periods, and local repair of clouded areas using cloud-free image information;

[0008] S4, quantitative correction based on cloud optical thickness, uses a cloud optical thickness model according to cloud type and makes gain correction to brightness at the pixel level;

[0009] S5. Geospatial grid maturity evaluation, which combines high-quality remote sensing data after cloud interference compensation with other multi-source big data indicators to obtain the final maturity of each grid cell.

[0010] In a preferred embodiment of the present invention, the prior detection based on meteorological data includes:

[0011] Obtain cloud data from meteorological station data and satellite cloud images, including cloud base height H base , cloud top height H top and cloud optical thickness τ(x,y,t);

[0012] Generate a cloud mask M(x,y,t) for the cloud-covered area, which is defined as follows:

[0013]

[0014] where x and y are geographical coordinates and t is time;

[0015] M(x,y,t) is the cloud mask, which is used to identify whether a pixel is covered by clouds;

[0016] When initially processing the original night light image I orig (x,y,t) obtained at night, mark the pixel values in the cloud-covered area as invalid values:

[0017] I masked (x,y,t) = I orig (x,y,t) · M(x,y,t);

[0018] where I masked (x,y,t) is the night light image after mask processing, which removes the information of the cloud-covered area;

[0019] I orig (x,y,t) is the original night light image, which is the original data before cloud interference correction;

[0020] In the subsequent correction and compensation process, only perform "cloud interference correction" on the pixels covered by clouds, and retain the original brightness information for the cloud-free area.

[0021] In a preferred embodiment of the present invention, the quantitative correction based on cloud optical thickness includes:

[0022] Establish a set of preferred rules to prioritize historical images according to time sequence and cloud coverage, and dynamically select the optimal reference image:

[0023] I ref (x,y,t′) = argmin t′∈P(cloudCoverage(t′)+γ·|t - t′|);

[0024] wherein, I ref (x, y, t′) is the night brightness value of the selected historical reference image;

[0025] P is the time index of the historical image set;

[0026] argmin t′∈P (*) means taking the t′ that makes the expression in the brackets the smallest as the optimal choice, that is, selecting the historical image with the smallest cloud coverage and a relatively small time difference;

[0027] cloudCoverage(t′) represents the cloud coverage at time t′, and γ is the weight coefficient of the time difference, controlling the influence degree of the time distance;

[0028] |t - t′| is the time difference, representing the time interval between the current time t and the historical image time t′. Images with a large time interval are not suitable as references due to large environmental changes.

[0029] In a preferred embodiment of the present invention, the multi - temporal contrast correction includes:

[0030] At the same coordinate (x, y) and in a similar time period, select the night brightness value I ref (x, y, t′) of the historical reference image, where t′ is the historical time;

[0031] If a certain pixel in the night image at time t is covered by clouds, the reference brightness in the historical image can be used for partial repair:

[0032]

[0033] wherein, I repaired (x, y, t) is the brightness value of the corrected night image, which is used to compensate the brightness of the pixel point when the brightness at the coordinate (x, y) is damaged due to cloud cover;

[0034] η is the correction coefficient, which is dynamically set according to the degree of cloud cover and the time interval, and the value range is η ∈ (0, 1);

[0035] When the cloud cover is relatively serious, η increases. If the cloud cover is not serious or the time interval is relatively long, then η decreases;

[0036] For the night brightness images obtained at different times or different angles in the same observation area, after spatially aligning these images, calculate the brightness differences of multiple scenarios:

[0037] ΔI k (x, y)=Ik (x,y)-I0(x,y),k∈{1,2,3,...,n};

[0038] Among them, ΔI k (x,y) is the brightness difference compared with the reference image under different observation conditions;

[0039] I k (x, y) is the nighttime brightness value of the kth set of observation images, which is data obtained from multiple satellites or multiple orbits;

[0040] I0(x,y) is the nighttime brightness value of the reference image, with the image with the least cloud cover or the best quality as the reference;

[0041] k is used to index observation images at different times or angles. A total of n groups of images can be used for cross-comparison;

[0042] The differences were analyzed by statistical methods, outliers were removed using Z-score, and the cloud cover was supplemented by spatial interpolation, which can further improve the correction accuracy of the cloud interference area.

[0043] In a preferred embodiment of the present invention, the quantitative correction based on cloud optical thickness includes:

[0044] In the case of penetrable clouds, the transmittance T(x,y,t) is estimated based on the cloud optical thickness τ(x,y,t):

[0045] T(x,y,t)=e -τ(x,y,t) ;

[0046] The cloud optical thickness τ(x, y, t) can be estimated by meteorological data. According to the cloud type, the cloud base height H base 、Cloud top height H top Fit using a physical model:

[0047] τ(x,y,t)=α·(H top -H base )+β·cloudTypeFactor;

[0048] Among them, α and β are unknown coefficients;

[0049] cloudTypeFactor represents the influence factor of different cloud types on optical thickness;

[0050] For areas covered by clouds where some visible / detectable brightness information is still present, a quantitative gain can be calculated based on the transmittance:

[0051]

[0052] Among them, θ is the threshold, the set minimum transmittance limit;

[0053] ω is an adjustable parameter used to control the brightness gain amplitude;

[0054] Only when the transmittance is such that 1≥T(x,y,t)>θ, can part of the light penetrate the cloud layer, and only then can the attenuation of the brightness caused by the cloud layer be compensated through this mode;

[0055] When τ(x,y,t) is too large resulting in θ≥T(x,y,t)>0, a masking process is performed on the cloud-covered area, and the previously described multi-temporal contrast correction method is used to fill in or infer the brightness value.

[0056] In a preferred embodiment of the present invention, the evaluation of the geospatial grid maturity includes:

[0057] Fusing the high-quality remote sensing image metrics obtained after cloud interference detection and compensation with multi-source big data metrics such as economy, infrastructure, and environment, and using the weighted sum of multi-dimensional data to obtain the comprehensive maturity score S maturity (g i ):

[0058]

[0059] Among them, S maturity (g i ) represents the maturity score of the grid cell g i ;

[0060] A is the weight of the night brightness, indicating the contribution degree of the night brightness index after cloud interference detection and compensation in the maturity calculation;

[0061] B is the weight of the economic activity level, indicating the contribution degree of the regional economic development situation in the maturity calculation;

[0062] C is the weight of the infrastructure completeness, reflecting the contribution of the infrastructure to the maturity calculation;

[0063] D is the weight of the environmental index, indicating the influence weight of the environmental factor in the maturity calculation;

[0064] represents the night brightness value obtained by extracting or aggregating the grid cell g i after cloud interference detection and compensation;

[0065] E econ (g i ) is an economic activity index, including quantitative indexes such as GDP, commercial facility density, traffic flow, electricity consumption, and night light intensity that can reflect the regional economic and social development level;

[0066] E infra (g i ) is the infrastructure completeness index, which is used to measure the urbanization level and public service supporting capacity of grid cells;

[0067] E env (g i ) is the environmental index;

[0068] This score is visualized on the big data platform to assist subsequent urban optimization decisions.

[0069] The present invention also discloses a computer system, including:

[0070] A processor;

[0071] A memory for storing processor-executable instructions;

[0072] Wherein, when the processor is configured to execute the executable instructions, it implements the above-mentioned method for evaluating and optimizing the maturity of geospatial grids based on big data.

[0073] The present invention also discloses a computer system, including:

[0074] A memory having a computer program stored thereon;

[0075] A processor for executing the program in the memory to implement the above-mentioned method for evaluating and optimizing the maturity of geospatial grids based on big data.

[0076] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are:

[0077] Precise cloud cover identification and multi-temporal correction: Based on meteorological data or satellite cloud images, prior information on cloud cover is obtained, and the cloud boundary is identified and the brightness is repaired for night images by combining the multi-temporal difference method, which can effectively reduce the optical information loss and misjudgment caused by cloud occlusion.

[0078] Multi-source data fusion and continuous cloud cover compensation: When there is continuous cloud cover for multiple days, high-resolution microwave radar remote sensing data or ground sensor data is used for auxiliary compensation, avoiding the defect that the "cloudless image" depending only on the same time phase or a similar time period cannot be repaired, and improving the feasibility of night observation under extreme weather conditions.

[0079] Quantitative gain correction based on cloud optical thickness: The cloud optical thickness model is used to estimate the transmittance of clouds, and differential gain correction is performed according to elements such as cloud base height, cloud top height, and cloud type, which can more precisely restore the actual brightness of the penetrable part of clouds and achieve adaptive processing for different cloud environments.

[0080] By dividing the target area into several grid cells, fusion analysis is performed on multi-source data (including satellite night light images, meteorological cloud images, ground sensor information, etc.) within each grid cell. On the one hand, cloud interference detection and compensation technology is used to improve the clarity of satellite images; on the other hand, big data algorithms are used to comprehensively evaluate economic, social, environmental and other indicators of each grid to form a geographical space grid maturity score. Finally, by combining temporal and spatial analysis, regional development planning and resource allocation are optimized and decision-making support is provided.

[0081] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] The above and / or additional aspects and advantages of the present invention will become apparent and be readily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0083] Figure 1 is a schematic diagram of the system of the present invention.

[0084] Figure 2 is an example of the visualized maturity. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0085] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described by referring to the drawings are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0086] As Figure 1 shown, the present invention discloses a method for optimizing the evaluation of geographical space grid maturity based on big data, including:

[0087] Based on the prior detection of meteorological data, and obtaining the information of meteorological stations or satellite cloud images in real time, masking the cloud-covered area in the night light image.

[0088] For the situation where the brightness of the cloud-covered area may be extremely low or disordered, the cloud edge is located by combining the meteorological cloud image.

[0089] The multi-temporal comparison method is used to cross-compare multi-scene night light images of the same area in a similar time period, and the cloud-covered area is locally repaired using the cloud-free image information.

[0090] If there is cloud interference for consecutive days, the light valuation of the corresponding area is obtained from high-resolution microwave radar remote sensing or ground-based sensor data.

[0091] Improve the quality of satellite images through cloud interference detection and compensation, thereby supporting the optimization of geospatial grid maturity evaluation.

[0092] And based on the quantitative correction of the cloud transmission model:

[0093] If it can partially penetrate the cloud layer, the cloud optical thickness model (estimating the actual attenuation based on cloud type, cloud base height, and cloud top height) is used to perform "gain correction" on the brightness at the pixel level.

[0094] Combined with the high-quality remote sensing image indicators obtained after cloud interference detection and compensation, multi-source big data is fused for geospatial grid maturity evaluation.

[0095] 1. Cloud cover detection and masking based on meteorological prior data

[0096] 1.1 Meteorological data reading and cloud map preprocessing:

[0097] Obtain cloud data from meteorological station data and satellite cloud maps, including cloud base height H base , cloud top height H top and cloud optical thickness τ(x, y, t);

[0098] Generate a cloud mask M(x, y, t) for the cloud-covered area, and its definition is as follows:

[0099]

[0100] where x and y are geographical coordinates, and t is time;

[0101] M(x, y, t) is the cloud mask, used to identify whether a pixel is covered by a cloud;

[0102] 1.2 Night image preprocessing based on mask information

[0103] When initially processing the original night light (brightness) image I orig (x, y, t) obtained at night, mark the pixel values in the cloud-covered area as invalid values:

[0104] I masked (x, y, t) = I orig (x, y, t) · M(x, y, t);

[0105] where I masked (x, y, t) is the night light (brightness) image after masking processing, removing the information of the cloud-covered area;

[0106] I orig (x, y, t) is the original night light (brightness) image, the original data before cloud interference correction;

[0107] In the subsequent correction and compensation process, only the pixels covering the cloud layer are subject to "cloud interference correction", while the original brightness information is retained for the cloud-free areas.

[0108] 2. Multi-temporal contrast correction method

[0109] 2.1 Selection of historical reference images

[0110] At the same coordinate (x, y) and in a similar time period, select the night brightness value I ref (x, y, t′) of the historical reference image, where t′ is the historical time.

[0111] 2.2 Cloud differential correction

[0112] If a certain pixel in the night image at time t is covered by clouds, partial repair can be performed using the reference brightness in the historical image:

[0113]

[0114] where, I repaired (x, y, t) is the brightness value of the corrected night image, which is used to compensate the brightness of this pixel point when the brightness at the coordinate (x, y) is damaged due to cloud cover;

[0115] η is the correction coefficient, which is dynamically set according to the degree of cloud cover and the time interval, and the value range is η ∈ (0, 1);

[0116] When the cloud cover is relatively serious, η increases. If the cloud cover is not serious or the time interval is relatively long, then η decreases.

[0117] 2.3 Cross-comparison of multiple images across scenes

[0118] For the night brightness images obtained for the same observation area at different times or different angles (multiple satellites, multiple observation orbits, etc.), after spatially aligning (registering) these images, calculate the brightness differences among multiple scenes:

[0119] ΔI k (x, y) = I k (x, y) - I0(x, y), k ∈ {1, 2, 3,....., n};

[0120] where, ΔI k (x, y) is the brightness difference compared with the reference image under different observation conditions (including time or angle);

[0121] I k (x, y) is the night brightness value of the kth group of observation images, which comes from the data obtained by multiple satellites or multiple orbits;

[0122] I0(x,y) is the nighttime brightness value of the reference image, with the image having the least cloud cover or the best quality being used as the reference;

[0123] k is used to index the observed images at different times or different angles, and there are a total of n groups of images available for cross-comparison;

[0124] The differences are analyzed through statistical methods (including median, mean, or confidence interval), outliers are removed, and the cloud-covered parts are supplemented. This can further improve the correction accuracy of the cloud interference area.

[0125] 3. Quantitative Correction Based on Cloud Optical Thickness

[0126] 3.1 Cloud Optical Thickness and Transmittance Model

[0127] In scenarios where clouds can be penetrated (such as high cloud bases, thin clouds, etc.), the transmittance T(x,y,t) can be estimated based on the cloud optical thickness τ(x,y,t):

[0128] T(x,y,t) = e -τ(x,y,t) ;

[0129] Among them, the cloud optical thickness τ(x,y,t) can be estimated from meteorological data, and according to the cloud type, cloud base height H base , cloud top height H top is fitted using a physical model:

[0130] τ(x,y,t) = α·(H top -H base ) + β·cloudTypeFactor;

[0131] Among them, α and β are undetermined coefficients;

[0132] cloudTypeFactor represents the influence factor of different cloud types on the optical thickness (such as cirrus clouds, stratus clouds, cumulus clouds, etc.);

[0133] 3.2 Pixel-level "Gain Correction"

[0134] For the cloud-covered area where there is still some visible / detectable brightness information, quantitative gain can be performed based on the transmittance:

[0135]

[0136] Among them, θ is the threshold, the set minimum transmittance limit;

[0137] ω is an adjustable parameter used to control the brightness gain amplitude;

[0138] Only when the transmittance is such that 1≥T(x,y,t)>θ and part of the light can penetrate the cloud layer, can the attenuation of the cloud layer on the brightness be compensated by this mode;

[0139] When τ(x,y,t) is too large resulting in θ≥T(x,y,t)>0, a masking process is performed on the cloud-covered area, and the previously described multi-temporal contrast correction method is used to fill in or infer the brightness value.

[0140] 4. Dynamic Selection of Historical Reference Data

[0141] Establish a set of optimization rules to prioritize historical images according to time sequence and cloud coverage, and dynamically select the optimal reference image. This process can be described as:

[0142] I ref (x,y,t′)=argmin t′∈P (cloudCoverage(t′)+γ·|t - t′|);

[0143] where, I ref (x,y,t′) is the nighttime brightness value of the selected historical reference image;

[0144] P is the time index of the set of historical images;

[0145] argmin t′∈P (*) means taking the t′ that makes the expression in the parentheses the smallest as the optimal choice, that is, selecting the historical image with the smallest cloud coverage rate and a relatively small time difference;

[0146] cloudCoverage(t′) represents the cloud coverage rate at time t′, and γ is the weight coefficient of the time difference, which controls the influence degree of the time distance;

[0147] |t - t′| is the time difference, representing the time interval between the current time t and the time t′ of the historical image. Images with a relatively large time interval are not suitable as references due to significant environmental changes.

[0148] 5. Evaluation of Geospatial Grid Maturity

[0149] Fuse the high-quality remote sensing image indicators obtained after cloud interference detection and compensation with multi-source big data indicators such as economy, infrastructure, and environment, and use the weighted sum of multi-dimensional data to obtain the comprehensive maturity score S maturity (g i ):

[0150]

[0151] where, S maturity (g i ) represents the grid cell gi Maturity score;

[0152] A is the weight of night brightness, indicating the contribution degree of the night brightness index after cloud interference detection and compensation in the maturity calculation;

[0153] B is the weight of economic activity, indicating the contribution degree of the regional economic development situation in the maturity calculation;

[0154] C is the weight of infrastructure completeness, reflecting the contribution of infrastructure to the maturity calculation;

[0155] D is the weight of environmental indicators, indicating the influence weight of environmental factors in the maturity calculation;

[0156] Indicates the night brightness value obtained by extracting or aggregating the grid cell g i after cloud interference detection and compensation;

[0157] E econ (g i ) is an economic activity indicator, including quantitative indicators such as GDP, commercial facility density, traffic flow, electricity consumption, and night light intensity that can reflect the level of regional economic and social development;

[0158] E infra (g i ) is an infrastructure completeness indicator, used to measure the urbanization level and public service supporting capacity of grid cells;

[0159] E env (g i ) is an environmental indicator;

[0160] This score is visualized on the big data platform to assist in subsequent optimization decisions such as urban planning, regional resource allocation, or refined management.

[0161] 5.1 Score range and grade division

[0162] By analyzing the numerical range of the comprehensive maturity score S maturity (g i ), combined with historical data and expert experience, the grading threshold is determined;

[0163] In the geospatial grid maturity evaluation described in the present invention, since different indicators (such as night light brightness, economic activity, infrastructure, environmental indicators) have different dimensions and value ranges, in order to facilitate subsequent comprehensive calculation or comparison, these indicators need to be "normalized" processed;

[0164] The comprehensive maturity score S maturity (g i)After mathematical transformation, it is mapped to a unified and controllable numerical interval, that is, normalized to the interval of [0, 100], and three levels of "low maturity", "medium maturity", and "high maturity" are divided within this interval to give an intuitive level determination for each grid cell;

[0165] 5.2 Visualization and Spatial Comparison

[0166] As Figure 2 shown, with the help of big data visualization tools, the maturity score or level of each grid cell is presented on the map in the form of colors or symbols.

[0167] Managers or decision-makers can clearly understand the maturity differences in different regions through the spatial heat map.

[0168] Through time series visualization, the change trend of scores in different time periods can also be shown to assist in evaluating the effectiveness of policy implementation or project promotion.

[0169] 5.3 Conduct Diagnostic Analysis in Combination with Specific Indicators

[0170] When it is found that the scores of some grid cells are relatively low or have decreased significantly, the weights and scores of specific indicators such as economy, infrastructure, and environment in the model can be further checked.

[0171] By comparing night lights, brightness after cloud compensation, and economic data, etc., analyze whether insufficient economic activities, lack of transportation networks, or environmental problems lead to the low comprehensive score.

[0172] This can help locate the root cause of the problem and provide targeted guidance for subsequent policy formulation or resource investment.

[0173] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the claims and their equivalents.

Claims

1. An optimization method for evaluating the maturity of a geospatial grid based on big data, characterized in that, include: S1, based on prior detection of meteorological data, obtains information from meteorological stations or satellite cloud images in real time and masks the cloud-covered area in the night light image; S2, dynamic selection of historical reference data, optimal selection for compensating for image history effects to improve image accuracy; S3, multi-temporal phase comparison correction, cross-comparison of multiple night light images of the same area in similar time periods, and local repair of clouded areas using cloud-free image information; S4, quantitative correction based on cloud optical thickness, uses a cloud optical thickness model according to cloud type and makes gain correction to brightness at the pixel level; S5, geospatial grid maturity evaluation, combines high-quality remote sensing data after cloud interference compensation with other multi-source big data indicators to obtain the final maturity of each grid cell.

2. An optimization method for evaluating the maturity of a geospatial grid based on big data according to claim 1, characterized in that The prior detection based on meteorological data includes: Obtain cloud data from meteorological station data and satellite cloud images, including cloud base height H base , cloud top height H top and cloud optical thickness τ(x, y, t); Generate a cloud mask M(x,y,t) for the cloud covered area, which is defined as follows: Where x and y are geographic coordinates, and t is time; M(x,y,t) is the cloud mask, which is used to identify whether the pixel is covered by clouds; For the original night light image I orig (x, y, t) obtained at night, when performing preliminary processing, mark the pixel values in the cloud-covered area as invalid values: I masked (x, y, t) = I orig (x, y, t) · M(x, y, t); Among them, I masked (x, y, t) is the night light image after masking processing, removing the information of cloud-covered areas; I orig (x, y, t) is the original night luminous image, the original data before cloud interference correction; In the subsequent correction and compensation process, cloud interference correction is only performed on pixels covered by clouds, while the original brightness information is retained for cloud-free areas.

3. An optimization method for evaluating the maturity of a geospatial grid based on big data according to claim 2, characterized in that, The dynamic selection of the historical reference data includes: Establish a set of optimization rules to prioritize historical images by time and cloud coverage, and dynamically select the optimal reference image: I ref (x,y,t′) = argmin t′∈P (cloudCoverage(t′) + γ·|t - t′|); Among them, I ref (x, y, t′) is the nighttime brightness value of the selected historical reference image; P is the time index of the historical image set; argmin t′∈P (*) means to select the t' that minimizes the expression within the parentheses as the optimal choice, that is, to select the historical image with the smallest cloud coverage and a relatively small time difference; cloudCoverage(t′) represents the cloud coverage at time t′, γ is the weight coefficient of time difference, which controls the influence of time distance; |tt′| is the time difference, which represents the time interval between the current time t and the historical image time t′. Images with a large time interval are not suitable as references due to large environmental changes.

4. An optimization method for evaluating the maturity of a geospatial grid based on big data according to claim 3, characterized in that, The multi-time phase contrast correction includes: Select the nighttime brightness value I of the historical reference image at the same coordinates (x, y) and in a similar time period. ref (x, y, t′), where t′ is the historical time. If a pixel in the night image at time t is covered by clouds, it can be partially repaired using the reference brightness in the historical image: Among them, I repaired (x, y, t) is the corrected night image brightness value, which is used to compensate the brightness of the pixel when the brightness of the coordinate (x, y) is damaged due to cloud occlusion. η is the correction coefficient, which is dynamically set according to the degree of cloud coverage and time interval, and the value range is η∈(0,1); When the cloud cover is severe, η increases; if the cloud cover is not severe or the time interval is long, η decreases; For the nighttime brightness images of the same observation area acquired at different times or angles, these images are spatially aligned and the brightness differences of multiple scenes are calculated: ΔI k (x,y) = I k (x,y) - I0(x,y), k ∈ {1, 2, 3,....., n}; Among them, ΔI k (x, y) is the brightness difference compared with the reference image under different observation conditions; I k (x, y) is the night brightness value of the k-th group of observed images, which is data obtained from multiple satellites or multiple orbits; I0(x,y) is the nighttime brightness value of the reference image, with the image with the least cloud cover or the best quality as the reference; k is used to index observation images at different times or angles. A total of n groups of images can be used for cross-comparison; The differences were analyzed by statistical methods, outliers were removed using Z-score, and the cloud cover was supplemented by spatial interpolation, which can further improve the correction accuracy of the cloud interference area.

5. The optimization method for evaluating the maturity of a geospatial grid based on big data according to claim 2, characterized in that, The quantitative correction based on cloud optical thickness includes: In the case of penetrable clouds, the transmittance T(x,y,t) is estimated based on the cloud optical thickness τ(x,y,t): T(x, y, t) = e -τ(x,y,t) ; Among them, the cloud optical thickness τ(x, y, t) can be estimated from meteorological data. According to the cloud type, cloud base height H base , cloud top height H top is fitted using a physical model: τ(x,y,t) = α·(H top - H base ) + β·cloudTypeFactor; Among them, α and β are empirical coefficients, which are determined based on historical meteorological data; The cloudTypeFactor represents the influence factor of different cloud types on cloud optical thickness; For the part of the cloud-covered area where there is still some visible / detectable brightness information, quantitative gain can be performed according to the transmittance: where θ is the threshold, the set minimum transmittance limit; ω is an adjustable parameter used to control the brightness gain amplitude, and its value range is ω ∈ (-1, 0); Only when the transmittance is 1 ≥ T(x, y, t) > θ, some light can penetrate the cloud layer, and this mode can be used to compensate for the attenuation of the cloud layer on the brightness; When τ(x, y, t) is too large resulting in θ ≥ T(x, y, t) > 0, a masking process is performed on the cloud-covered area, and a multi-temporal comparison correction method is used to fill or infer the brightness value.

6. The optimized method for evaluating the maturity of a geospatial grid based on big data according to claim 1, characterized in that, The geospatial grid maturity evaluation includes: Fuse the high-quality remote sensing image metrics obtained after cloud interference detection and compensation with multi-source big data metrics such as economy, infrastructure, and environment, and use the weighted sum of multi-dimensional data to obtain the comprehensive maturity score S maturity (g i ): Among them, S maturity (g i ) represents the maturity score of the grid cell g i ; A is the weight of the night brightness, representing the contribution degree of the night brightness index after cloud interference detection and compensation in the maturity calculation; B is the weight of the economic activity level, representing the contribution degree of the regional economic development situation in the maturity calculation; C is the weight of the infrastructure completeness, reflecting the contribution of the infrastructure to the maturity calculation; D is the weight of the environmental index, representing the influence weight of environmental factors in the maturity calculation; Indicates the night brightness value obtained by extracting or aggregating the grid cell g i after cloud interference detection and compensation; E econ (g i ) is an economic activity indicator, including quantitative indicators such as GDP, commercial facility density, traffic flow, electricity consumption, and night light intensity that can reflect the level of regional economic and social development; E infra (g i ) is an infrastructure completeness index used to measure the urbanization level and public service supporting capacity of grid cells; E env (g i ) is an environmental indicator; This score is visualized on the big data platform to assist subsequent urban optimization decisions.

7. A computer system, characterized in that, Including: A processor; A memory for storing the executable instructions of the processor; wherein, when the processor is configured to execute the executable instructions, it implements an optimization method for geospatial grid maturity evaluation based on big data according to any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, Including: A memory with a computer program stored thereon; A processor for executing the program in the memory to implement an optimization method for geospatial grid maturity evaluation based on big data according to any one of claims 1 to 6.

Images