A multi-source night light data correction fusion method, a terminal device and a medium

Abstract

The present application relates to a kind of multi-source night light data correction fusion method, terminal equipment and medium, the method includes: S1: obtaining the DMSP-OLS and NPP-VIIRS night light data to be fused;S2: pre-processing is carried out to night light data;S3: selecting invariant target area, using invariant target method, the DMSP-OLS night light data after pre-processing is successively carried out between sensors mutual correction, image intra-annual fusion and image inter-annual continuity correction;S4: DMSP-OLS and NPP-VIIRS night light data are mutually corrected;S5: the data after step S4 mutual correction is carried out image inter-annual continuity correction.The present application improves the continuity and comparability of long time period time series multi-source night light data set.

Description

Technical Field

[0001] This invention relates to the field of data fusion, and in particular to a method, terminal device and medium for correcting and fusing multi-source nighttime light data. Background Technology

[0002] Currently, there are two commonly used sensors internationally for detecting nighttime light intensity on the Earth's surface: one is the Operational Linescan System nighttime stable light data (DMSP-OLS) of the Defense Meteorological Satellite Program, and the other is the Visible Infrared Imaging Radiometer Suite Day-night Band (NPP-VIIRS DNB) on the Suomi National Polar-Orbiting Partnership (S-NPP) satellite. DMSP-OLS can detect urban lights, lightning, and fires on the Earth's surface at night, creating a striking contrast with the dark rural background in the image data. However, DMSP-OLS was the first to acquire nighttime light image data and suffers from limitations such as coarse radiometric measurement accuracy, low spatial resolution, and a lack of onboard calibration. Furthermore, due to the lack of onboard calibration, variable atmospheric conditions, and sensor degradation, the annual dynamic variations of DMSP-OLS time-series nighttime light data cannot be directly compared. Later, the new generation of high-resolution NPP-VIIRS nighttime light remote sensing data was more readily available and had radiometric calibration features, overcoming some limitations and shortcomings of DMSP-OLS data. NPP-VIIRS nighttime light remote sensing data has higher satellite-borne calibrated radiometric measurement accuracy, ensuring the ability to detect nighttime lights with extremely low brightness.

[0003] Currently, the fourth edition of DMSP-OLS non-radiometrically calibrated average nighttime light intensity imagery data is historical archive data. This data currently includes 34 image periods, spanning from 1992 to 2013, with DN values ​​ranging from 0 to 63 for each period. This stable nighttime light remote sensing data is generally used for quantitative inversion studies of urbanization monitoring. However, it lacks continuity and comparability between different periods and suffers from data saturation. In contrast, NPP-VIIRS nighttime light remote sensing data is available from 2012 to the present, exhibiting better continuity. However, DMSP-OLS is long-term archived historical data, while NPP-VIIRS data has shorter annual periods, and there is little overlap between the two. Summary of the Invention

[0004] To address the aforementioned issues, this invention proposes a multi-source nighttime light data correction and fusion method, terminal equipment, and medium.

[0005] The specific plan is as follows:

[0006] A method for correcting and fusing multi-source nighttime light data includes the following steps:

[0007] S1: Acquire the nighttime light data of DMSP-OLS and NPP-VIIRS to be fused;

[0008] S2: Preprocess the nighttime light data of DMSP-OLS and NPP-VIIRS, including image reprojection, resampling and cropping;

[0009] S3: Select an invariant target area and use the invariant target method to sequentially perform inter-sensor mutual correction, intra-year image fusion, and inter-year image continuity correction on the preprocessed DMSP-OLS nighttime light data;

[0010] S4: The following model is used to cross-calibrate the nighttime light data of DMSP-OLS and NPP-VIIRS:

[0011] DN r =α×(LgVIIRS) 2 +β×(LgVIIRS)+γ

[0012] Among them, DN r represents the value of the corrected nighttime light image data, VIIRS represents the value of the NPP-VIIRS nighttime light image data before correction, LgVIIRS represents the logarithmic transformation of VIIRS, and α, β, and γ are regression parameters in the quadratic equation regression model.

[0013] S5: Perform interannual continuity correction on the data after mutual correction in step S4.

[0014] Furthermore, the specific preprocessing steps include: firstly, reprojecting the nighttime light data into an Albers equal-area conic projection, and then resampling it to 1km using a bilinear fitting method; secondly, obtaining nighttime light data with a spatial resolution of 1km through cropping.

[0015] Furthermore, the quadratic regression model used for mutual calibration between sensors is as follows:

[0016] DN c =a×DN 2 +b×DN+c

[0017] Among them, DN and DN c...

[0018] Furthermore, the model used for image fusion within the year is as follows:

[0019]

[0020] in, These represent the DN values ​​of the i-th pixel in the nighttime light data acquired by two different sensors, a and b, after mutual calibration in year n; DN (n,i) This represents the DN value of the i-th pixel in the nighttime light data of the nth year after correction.

[0021] Furthermore, the model used for interannual continuity correction of images is as follows:

[0022]

[0023] Among them, DN (n-1,i) DN (n,i) DN (n+1,i) DN' represents the DN value of the i-th pixel in the nighttime light data after image fusion in years n-1, n, and n+1, respectively. (n,i) This represents the DN value of the i-th pixel in the nighttime light data of the nth year after interannual continuity correction.

[0024] A multi-source nighttime light data correction and fusion terminal device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the method described above in the embodiments of the present invention.

[0025] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method described above in the embodiments of the present invention.

[0026] This invention employs the above-described technical solution to resolve the incomparability issues between different years' data in the DMSP-OLS nighttime light image dataset, as well as the incomparability between the DMSP-OLS nighttime light image dataset and the NPP-VIIRS nighttime light image dataset. This invention improves the continuity and comparability of long-term time-series multi-source nighttime light datasets, expands the application scope of nighttime light remote sensing data, and enhances the understanding of long-term urbanization processes and their impact on the ecological environment. Attached Figure Description

[0027] Figure 1 The diagram shown is a flowchart of Embodiment 1 of the present invention.

[0028] Figure 2The figure shown is a graph illustrating the changes in the average DN value and total pixel value of the DMSP-OLS nighttime light data in this embodiment.

[0029] Figure 3 The diagram shown is a statistical analysis result of the DN values ​​of the corrected DMSP-OLS time series nighttime light data in the study area in this embodiment.

[0030] Figure 4 The diagram shown is a schematic representation of the statistical analysis results of the DN values ​​of the corrected NPP-VIIRS time series nighttime light data in the study area in this embodiment.

[0031] Figure 5 The figure shows the average and total number of DN values ​​of the time series nighttime light data after correction for the two types of nighttime light data in the study area in this embodiment. Detailed Implementation

[0032] To further illustrate the various embodiments, the present invention provides accompanying drawings. These drawings are part of the disclosure of the present invention, primarily used to illustrate the embodiments, and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementations and the advantages of the present invention.

[0033] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0034] Example 1:

[0035] This invention provides a method for correcting and fusing multi-source nighttime light data, such as... Figure 1 The diagram shown is a flowchart of the multi-source nighttime light data correction and fusion method according to an embodiment of the present invention. The method includes the following steps:

[0036] S1: Acquire the nighttime light data from DMSP-OLS and NPP-VIIRS to be fused.

[0037] S2: Preprocess the nighttime light data of DMSP-OLS and NPP-VIIRS.

[0038] DMSP-OLS and NPP-VIIRS nighttime light data contain stable light data of towns and other types. After rigorous processing, the transient brightness of events such as fires and the influence of factors such as sunlight, moonlight, clouds, and auroras have been removed. However, due to the incomparability of data from different years in time series datasets, they cannot be directly used to extract information on urban built-up areas. The specific limitations are mainly reflected in four aspects: (1) the average intensity values ​​of nighttime light data acquired by the two satellites in the same year are different; (2) the average intensity values ​​of nighttime light data acquired by the same satellite in different years are abnormally fluctuating; (3) the number of pixels in nighttime light data acquired by the two satellites in the same year is different; and (4) the number of pixels in nighttime light data acquired by the same satellite in different years is abnormally reduced. Therefore, before carrying out research, it is necessary to perform relative radiometric calibration, intra-year data correction, and inter-year series correction on the nighttime light dataset to improve the continuity and comparability of the time series nighttime light dataset.

[0039] In this embodiment, the nighttime light remote sensing data is first preprocessed by image reprojection, resampling, and cropping. Specifically, the nighttime light remote sensing images from 1992 to 2012 (21 years) are first reprojected into the Albers Conical Equal Area Projection, and then resampled to 1 km using a bilinear fitting method. Next, cropping is performed to obtain non-radiatively calibrated nighttime light remote sensing data for the entire 21 years (1992-2012) of China with a spatial resolution of 1 km.

[0040] S3: Select an invariant target area and use the invariant target method to sequentially perform inter-sensor mutual correction, intra-year image fusion, and inter-year image continuity correction on the preprocessed DMSP-OLS nighttime light data.

[0041] Statistical analysis was conducted on the socio-economic statistics of major cities in China from 1992 to 2012, including GDP, urban population, and urban built-up area, to determine the invariant target area. In this embodiment, the invariant target area should meet the following conditions: (1) the major socio-economic statistics show little change during these 20 years, indicating that the city has been in a relatively stable development trend during this period; (2) the stable nighttime light data has a wide range of DN values ​​from low to high, ensuring the accuracy of the nighttime light data mutual correction model; (3) the nighttime light images acquired by the F12 satellite in 1992 and the nighttime light images acquired by the F18 satellite in 2012 have a good linear correlation, indicating that the nighttime light intensity of the city is relatively stable during this period. In this embodiment, after determining the invariant target area, the corresponding areas of the 33-period DMSP-OLS stable nighttime light image data and the F162006 radiometric calibrated DMSP-OLS nighttime light image data are determined as the image data to be corrected and the reference image data, respectively.

[0042] (1) Mutual calibration between sensors

[0043] Abnormal fluctuations in the DN values ​​of multi-sensor image pixels are the main cause of discontinuities in nighttime light imagery data. To improve the continuity of nighttime light data, inter-sensor cross-calibration is necessary. To effectively reduce the differences in DN values ​​between nighttime light data, a specific method for cross-calibration of global nighttime light remote sensing data based on a constant target area is proposed: A reference area is determined and a regression model is constructed; the reference area should be selected based on small interannual variations in DN values. Regression analysis is then performed between image data from other years and the DN values ​​of the determined calibration area in the reference data. Finally, the constructed regression model is used to calibrate the long-term series dataset.

[0044] Due to the significant differences in urban development levels between different regions, the actual development status of cities is taken into consideration. (1) First, after analyzing the GDP and urban built-up area data of major cities, cities with relatively stable socio-economic development are selected as reference areas; (2) Second, data from the F16 satellite in 2007 are selected as reference datasets; (3) The DN values ​​of nighttime light remote sensing data of reference areas in other years are analyzed by a quadratic regression model with the DN values ​​of nighttime light remote sensing data of the F16 reference area in 2007. The quadratic regression model is shown in Formula 1; (4) Finally, the regression parameters are obtained by using the constructed quadratic regression model, and the long-term series DMSP-OLS nighttime light remote sensing dataset is mutually corrected.

[0045] DN c =a×DN 2 +b×DN+c (1)

[0046] In the formula, DN, DNc , respectively, are the image grayscale values ​​before and after correction; a, b, and c all represent quadratic regression parameters.

[0047] In this embodiment, a quadratic regression model analysis was performed on the long-term DMSP-OLS nighttime light remote sensing data of the reference area to obtain its regression determination coefficient R. 2 All values ​​were above 0.83, and their regression determination coefficients R0 were obtained through linear regression model analysis. 2 All values ​​were above 0.82, indicating that the regression models were of good accuracy. Therefore, this embodiment uses the parameters of the quadratic regression model to perform mutual correction on DMSP-OLS nighttime light data within the China region from 1992 to 2012. The parameters of the quadratic regression model and the linear regression model are shown in Table 1.

[0048] Table 1

[0049]

[0050]

[0051] After inter-sensor cross-calibration, the long-term DMSP-OLS nighttime light datasets are comparable, while reducing the saturation of pixel DN values ​​in each period of nighttime light data. However, the discontinuity problem of the nighttime light datasets after cross-calibration remains unresolved, manifested in the existence of nighttime light data from multiple sensors acquired in the same year, and abnormal fluctuations in pixel DN values ​​for the same area in nighttime light datasets from different years acquired by multiple sensors. Therefore, intra-year image fusion calibration and inter-year image continuity calibration are still needed for the DMSP-OLS nighttime light image datasets.

[0052] (2) Image fusion within the year

[0053] Due to inherent differences between different satellite sensors, and the various factors that influence the acquisition of nighttime light data, discrepancies exist between nighttime light data acquired by multiple sensors for the same year. While mutual correction based on the invariant target correction method reduces these discrepancies, it does not completely eliminate them. Nighttime light data from two periods within the same year still exhibit differences. To fully utilize the independent acquisition of nighttime light and shadow data from multiple sensors for the same year and address pixel differences in the same region within the nighttime light data from multiple sensors for the same year, intra-year image fusion is necessary. Therefore, in this embodiment, formula (2) is used to perform intra-year image fusion on the nighttime light data from multiple sensors after mutual correction for the same year, forming a unique nighttime light image dataset for each year.

[0054]

[0055] In the formula, These represent the DN values ​​of the i-th pixel in the nighttime light data acquired by two different sensors, a and b, after mutual calibration in year n; DN (n,i) This represents the DN value of the i-th pixel in the nighttime light data of year n after correction; n = 1994, 1997, 1998, ..., 2007.

[0056] (3) Interannual continuity correction of images

[0057] Based on the accelerating urbanization process in China, it can be assumed that urban area luminance pixels acquired in earlier nighttime light imagery data will not disappear in later nighttime light imagery data. Studies show that luminance pixels with higher DN values ​​in nighttime light imagery are more likely to be extracted as urban areas. Therefore, the DN value of luminance pixels in earlier nighttime light imagery should not be greater than the DN value of luminance pixels in the same area in later nighttime light imagery data. In this context, in the long-term DMSP-OLS nighttime light imagery dataset, if a pixel exists in earlier nighttime light imagery but disappears in the same area in later nighttime light imagery, or if a pixel has a larger DN value in earlier nighttime light imagery data than in later nighttime light imagery data, this pixel data can be considered unstable, meaning that the nighttime light imagery dataset exhibits pixel fluctuations across different years.

[0058] Since mutual correction and intra-year fusion correction of nighttime light data sets have not resolved the abnormal fluctuation of pixel DN values ​​between nighttime light data acquired by multiple sensors from different years, inter-year correction is needed to improve the continuity of nighttime light data sets. The basis for inter-year correction is as follows: (1) When the pixel DN value in the later nighttime light data is 0, the pixel DN value at the same position in the earlier nighttime light data should also be 0; (2) When the pixel DN value in the later nighttime light data is not 0, the pixel DN value in the earlier nighttime light data should not be greater than the pixel DN value at the same position in the later nighttime light data. The specific correction formula is shown in Formula 3.

[0059]

[0060] In the formula, DN (n-1,i) DN (n,i) DN (n+1,i) DN' represents the DN value of the i-th pixel in the nighttime light data after image fusion in years n-1, n, and n+1, respectively. (n,i) This represents the DN value of the i-th pixel in the nighttime light data of the nth year after interannual continuity correction; n = 1992, 1993, ..., 2013.

[0061] In this embodiment, after correction using the three correction methods described above, regression fitting analysis was performed on the DMSP-OLS nighttime light data from satellites of various time series from 1992 to 2013 to obtain the regression determination coefficient R. 2 All values ​​were above 0.83, indicating good accuracy of the regression model.

[0062] To verify the rationality of the correction method for nighttime light remote sensing image data from 1992 to 2013, this embodiment employs both qualitative and quantitative methods for verification. The qualitative method uses visual interpretation, comparing the changes in pixel DN values ​​between nighttime light remote sensing data of the same year before and after correction to verify whether the corrected nighttime light remote sensing data resolved the problems present before correction. This method is simple and intuitive, but highly subjective. Therefore, a quantitative analysis method is also needed to verify the corrected nighttime light data, making the verification results more accurate, objective, and reliable. Quantitative verification compares the changes over time in the total DN value (TDN) and the total number of luminous pixels (TLP) of the corresponding nighttime light remote sensing data before and after correction. The formula for calculating the total DN value (TDN) of nighttime light remote sensing data is shown in Formula 4.

[0063]

[0064] In the formula, DN i C represents the DN value of the i-th pixel in the nighttime light image; i This indicates the number of pixels in the nighttime light image.

[0065] This embodiment also verifies the corrected nighttime light image data. First, it examines the changes in the average DN value and total pixel value of the DMSP-OLS nighttime light data from 1992-2013 before correction, such as... Figure 2 As shown, it can be concluded that there are significant differences in the nighttime light data acquired by different sensors for the same year before correction, and that the nighttime light data acquired by the same sensor for different years exhibit discontinuities such as abnormal reduction in pixels. Secondly, the changes in the average DN value and total pixel value of the DMSP-OLS nighttime light data from 1992 to 2013 after correction are shown in the following figures. Figure 3 As shown, the corrected time series nighttime light remote sensing data has a unique value for each year, and there is no abnormal decrease in the number of pixels in the nighttime light data from different years. The average DN value and the total number of DN values ​​of the nighttime light remote sensing data both show a stable increasing trend. Therefore, the corrected DMSP-OLS time series nighttime light data has good consistency and the results are relatively accurate.

[0066] S4: Perform sensor cross-calibration on nighttime light data of DMSP-OLS and NPP-VIIRS.

[0067] Because the data sources for DMSP-OLS nighttime light data and NPP-VIIRS nighttime light data are different, this embodiment also includes registration and correction of the DMSP-OLS nighttime light data and NPP-VIIRS nighttime light data.

[0068] First, in this embodiment, quadratic and logarithmic regression analyses were performed on the 2013-2020 NPP-VIIRS long-term nighttime light data and the 2012 DMSP-OLS nighttime light data of the reference area. The accuracy of the regression models is shown in Tables 2 and 3, indicating that the accuracy of the regression models is good. Second, in this embodiment, the regression model parameters were also used to cross-calibrate the NPP-VIIRS nighttime light remote sensing data of the reference area from 2013 to 2020. For ease of description, the calibrated NPP-VIIRS nighttime light data is referred to as "DMSP-OLS-like" nighttime light data in this embodiment.

[0069] In this embodiment, the calibration between the two types of sensors, DMSP-OLS nighttime light data and NPP-VIIRS nighttime light data, is expressed by the formula shown in Equation 5.

[0070] DN r =α×(LgVIIRS) 2 +β×(LgVIIRS)+γ (5)

[0071] In the formula, DN r α represents the value of the corrected nighttime light image data, VIIRS represents the value of the NPP-VIIRS nighttime light image data before correction, LgVIIRS represents the logarithmic transformation of VIIRS, and α, β, and γ are regression parameters in the quadratic equation regression model.

[0072] Table 2

[0073]

[0074]

[0075] S3: Perform interannual continuity correction on the data (DMSP-OLS-like nighttime light data) after mutual correction in step S4.

[0076] The basis for the interannual continuity correction of DMSP-OLS nighttime light data is as follows: (1) When the DN value of a pixel in the later-corrected DMSP-OLS nighttime light data is 0, the DN value of a pixel at the same position in the earlier-corrected nighttime light data should also be 0; (2) When the DN value of a pixel in the later-corrected DMSP-OLS nighttime light data is not 0, the DN value of a pixel in the earlier-corrected DMSP-OLS nighttime light data should not be greater than the DN value of a pixel at the same position in the later-corrected DMSP-OLS nighttime light data. The specific correction formula is shown in Formula 6.

[0077]

[0078] In the formula, DN (n-1,i) DN (n,i) DN (n+1,i) These represent the DN values ​​of the i-th pixel in the nighttime light images acquired by multiple sensors in years n-1, n, and n+1, respectively, after mutual correction and inter-sensor correction; n = 2013, 2014, ..., 2020.

[0079] Table 3

[0080]

[0081]

[0082] This embodiment performs a quadratic regression fitting analysis on the logarithmized nighttime light remote sensing data of NPP-VIIRS from 2013 to 2020 and the DMSP-OLS nighttime light remote sensing data from 2012, and obtains the regression determination coefficient R. 2 All values ​​were above 0.71, indicating good accuracy of the regression model.

[0083] To verify the correction and fusion results of long-term multi-source nighttime light data from DMSP-OLS and NPP-VIIRS, the rationality of the correction and fusion results and methods of NPP-VIIRS nighttime light images from 2013 to 2020 was first verified. In this embodiment, a quantitative method using the average value of nighttime light data and the total nighttime light brightness of the time-series data was used for evaluation and verification. The logarithmic quadratic regression model method proposed in this embodiment was used to correct the NPP-VIIRS nighttime light data. The quantitative evaluation results of the NPP-VIIRS corrected data are as follows: Figure 4 As shown, the corrected NPP-VIIRS nighttime light data from 2013 to 2020 shows a steady increase year by year, which is consistent with the socio-economic development of the study area. Therefore, the correction results of the NPP-VIIRS nighttime light remote sensing data are good.

[0084] A quantitative evaluation of the correction of time-series multi-source nighttime light data from DMSP-OLS and NPP-VIIRS was conducted, and the trends of TDN and TLP of luminance pixels in the corrected nighttime light image data of both systems over time were statistically analyzed. Based on the aforementioned correction of DMSP-OLS time-series nighttime light data and the correction methods for multi-source nighttime light data from DMSP-OLS and NPP-VIIRS, a series of correction processes were performed. Finally, the average DN and total DN values ​​of the corrected time-series nighttime light data from DMSP-OLS and NPP-VIIRS over a long period from 1992 to 2020 were obtained, as shown below. Figure 5 As shown; among them, the data from 1992 to 2012 are DMSP-OLS nighttime light remote sensing image data, and the data from 2013 to 2020 are "DMSP-OLS-like" nighttime light image data.

[0085] Figure 5 The results show that the total light intensity (TLP) and total density (TDN) of the corrected long-term series of nighttime light imagery data are increasing. The statistical difference in TLP gradually decreased between 2014 and 2020. TDN increased steadily between 1992 and 2006, increased sharply between 2007 and 2013, and then stabilized between 2014 and 2020. This is mainly because the nighttime light intensity in the urban center of the Guangdong-Hong Kong-Macao Greater Bay Area city cluster tends to stabilize in the mid-to-late stage of development. The nighttime light imagery data exhibits stable image data, indicating that the expansion of the city cluster is a continuous expansion and development from the central area to the surrounding areas. This can also be derived by observing the floating-point data of NPP-VIIRS nighttime light imagery. Furthermore, it can be clearly seen from point 5 that: after correction, each year's nighttime light data in the long-term multi-source nighttime light remote sensing data from 1992 to 2020 has a unique value; and there is no instance where the nighttime light data value of the following year is lower than that of the previous year, but rather shows a trend of increasing year by year and a stable growth, indicating that the correction and fusion results of the DMSP-OLS and NPP-VIIRS multi-source nighttime light remote sensing datasets are good.

[0086] After the above steps are performed to correct the DMSP-OLS and NPP-VIIRS data, they can be considered as a single data set for fusion.

[0087] This invention, through relative radiometric calibration, intra-year image correction, and inter-year sequence correction of DMSP-OLS nighttime light image data, corrects NPP-VIIRS nighttime light image data to "DMSP-OLS-like" nighttime light data. This improves the continuity and comparability of long-term time-series multi-source nighttime light image datasets, expands the application scope of nighttime light remote sensing data, enhances the understanding of long-term time-series urbanization processes, human activities, and their impact on the ecological environment, and facilitates subsequent research using long-term multi-source nighttime light data.

[0088] The multi-source nighttime light image correction and fusion method proposed in this embodiment improves the continuity and comparability of long-term multi-source nighttime light data. The urbanization process is directly reflected in the transformation of urban land use / cover types, and changes in the brightness range of nighttime light remote sensing images can inversely reflect changes in urban land use / cover types. Since nighttime light remote sensing data is readily available, the correction process proposed in this embodiment is repeatable, and the corrected nighttime light image data exhibits continuity and stability. The DN value saturation phenomenon and spillover effect are also significantly reduced after correction, making the corrected nighttime light remote sensing data an important data source for studying urban land use / cover changes. Therefore, this embodiment can be applied to longer-term dynamic monitoring of the urbanization expansion and spatiotemporal evolution of large-scale urban agglomerations.

[0089] The method in this embodiment has the following key points:

[0090] Key Point 1: For time-series DMSP-OLS nighttime light remote sensing data, the "quadratic regression model method in the invariant region" is adopted to solve the problem of incomparability of time-series DMSP-OLS nighttime light datasets across different years. Specifically, it solves the following limitations of time-series DMSP-OLS: (1) differences in the average intensity values ​​of nighttime light data acquired by two satellites in the same year; (2) abnormal fluctuations in the average intensity values ​​of nighttime light data acquired by the same satellite in different years; (3) differences in the number of pixels in nighttime light data acquired by two satellites in the same year; and (4) an abnormal decrease in the number of pixels in nighttime light data acquired by the same satellite in different years.

[0091] Key Point Two: For the DMSP-OLS and NPP-VIIRS nighttime light remote sensing data, this invention proposes a "logarithmic quadratic term regression model method" for correction, which solves the incomparability problem between the DMSP-OLS nighttime light dataset and the NPP-VIIRS nighttime light dataset, and uses TDN to quantitatively evaluate the nighttime light after correction.

[0092] Key Point 3: Addressing the limitations of saturation in DMSP-OLS nighttime light remote sensing data. By employing the inter-sensor mutual calibration in the "quadratic regression model method in the invariant region" constructed in this invention, the saturation phenomenon in DMSP-OLS nighttime light remote sensing image data can be eliminated.

[0093] Key Point 4: To address the discontinuity of the "DMSP-OLS-like nighttime light data (data after mutual correction in step S4)," inter-year continuity correction of the "DMSP-OLS-like nighttime light data" is performed. The correction is based on the following: (1) When the DN value of a pixel in the later-corrected DMSP-OLS-like nighttime light data is 0, the DN value of a pixel at the same position in the earlier-corrected nighttime light data should also be 0; (2) When the DN value of a pixel in the later-corrected DMSP-OLS-like nighttime light data is not 0, the DN value of a pixel in the earlier-corrected DMSP-OLS-like nighttime light data should not be greater than the DN value of a pixel at the same position in the later-corrected DMSP-OLS-like nighttime light data, as shown in formula (6).

[0094] In summary, this embodiment improves the continuity and comparability of long-term time-series multi-source nighttime light datasets, providing important technical support for applied research on long-term nighttime light data. Furthermore, it significantly expands the application research of long-term multi-source nighttime light remote sensing data, enhancing our understanding of long-term time-series urbanization monitoring, human activities, and their impact on the ecological environment.

[0095] Example 2:

[0096] The present invention also provides a multi-source nighttime light data correction and fusion terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the method embodiment described above in Embodiment 1 of the present invention.

[0097] Furthermore, as an executable solution, the multi-source nighttime lighting data correction and fusion terminal device can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. The multi-source nighttime lighting data correction and fusion terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that the above-described structure of the multi-source nighttime lighting data correction and fusion terminal device is merely an example and does not constitute a limitation on the multi-source nighttime lighting data correction and fusion terminal device. It may include more or fewer components than described above, or combine certain components, or different components. For example, the multi-source nighttime lighting data correction and fusion terminal device may also include input / output devices, network access devices, buses, etc., and this embodiment of the invention does not limit this.

[0098] Furthermore, as an executable solution, the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the multi-source nighttime lighting data correction and fusion terminal equipment, connecting various parts of the entire multi-source nighttime lighting data correction and fusion terminal equipment via various interfaces and lines.

[0099] The memory can be used to store the computer programs and / or modules. The processor, by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory, realizes various functions of the multi-source nighttime light data correction and fusion terminal device. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0100] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method described in the embodiments of the present invention.

[0101] If the modules / units integrated in the multi-source nighttime light data correction and fusion terminal device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), and a software distribution medium, etc.

[0102] Although the invention has been specifically shown and described in conjunction with preferred embodiments, those skilled in the art should understand that various changes in form and detail may be made to the invention without departing from the spirit and scope of the invention as defined in the appended claims, all of which shall be within the scope of protection of the invention.

Claims

1. A multi-source night light data correction fusion method, characterized in that, The method comprises the following steps: S1: obtaining DMSP-OLS and NPP-VIIRS night light data to be fused; S2: preprocessing the DMSP-OLS and NPP-VIIRS night light data, wherein the preprocessing comprises image re-projection, re-sampling and clipping; the specific process of the preprocessing comprises: firstly, re-projecting the night light data into an Albers equal-area conic projection and re-sampling into 1km by using a bilinear fitting method; secondly, obtaining night light data with a spatial resolution of 1km by clipping; S3: selecting an invariant target region, and using an invariant target method to sequentially perform mutual correction between sensors, image inter-annual fusion and image inter-annual continuity correction on the preprocessed DMSP-OLS night light data; S4: using the following model to perform mutual correction between the DMSP-OLS and NPP-VIIRS night light data: DN r = a x (LgVIIRS) 2 + b x (LgVIIRS) + g wherein DN r represents the value of the corrected night light image data, VIIRS represents the value of the NPP-VIIRS night light image data before correction, LgVIIRS represents the logarithm of VIIRS, and a, β, and γ are regression parameters in the quadratic equation regression model. S5: performing image inter-annual continuity correction on the data after the mutual correction of step S4; The quadratic regression model used for mutual correction between sensors is: DN c = a x DN 2 + b x DN + c wherein DN and DN c respectively represent the image gray value before and after correction; a, b, c all represent quadratic regression parameters; The model used for image inter-annual fusion is: wherein, DNn,i and DNn,i represent the DN value of the i-th pixel of the night light data of the two different sensors a and b after mutual correction in the n-th year; DN (n,i) DNn,i represents the DN value of the i-th pixel of the night light data in the n-th year after correction. The model used for image inter-annual continuity correction is: DN (n-1,i) , DN (n,i) , DN (n+1,i) respectively represent the DN value of the i-th pixel of the night light data fused in the image year of the n-1th, nth and n+1th year, respectively, and DN' (n,i) represents the DN value of the i-th pixel of the night light data of the n-th year after the inter-annual continuity correction. 2.A multi-source night light data correction and fusion terminal device, characterized in that: The computer program is executed by the processor to implement the steps of the method according to any one of claims 1.

3. A computer readable storage medium storing a computer program, characterized in that: The computer program is executed by the processor to implement the steps of the method according to any one of claims 1.

Images