Satellite remote sensing image jitter detection and correction method and system

Abstract

The application provides a satellite remote sensing image tremor detection and correction method and system, and relates to the technical field of satellite remote sensing image processing. The method comprises the following steps: S1, selecting a ground object in a remote sensing image as an analysis object; S2, performing edge detection and extraction on the analysis object; S3, fitting the result of the edge detection; S4, identifying the tremor error of the remote sensing image according to the result of the fitting; S5, performing frequency analysis on the tremor error of the remote sensing image to identify the frequency characteristics of the tremor error; and S6, performing image correction based on the remote sensing image resampling method. The application can identify without relying on external measurement data through image information, and is convenient for popularization and application.

Description

Technical Field

[0001] This invention relates to the field of satellite remote sensing image processing technology, and more specifically, to a method and system for detecting and correcting flutter in satellite remote sensing images. Background Technology

[0002] Remote sensing imagery is the core product of remote sensing satellites, and its quality directly affects the satellite's operational effectiveness. During the satellite's Earth imaging process, micro-vibrations can cause image jitter, severely impacting the quality of subsequent remote sensing products. Therefore, it is necessary to conduct research on remote sensing image jitter detection and correction methods.

[0003] Chinese patent application CN107292839B discloses a TDI flutter image restoration method based on adaptive adjustment of image blocks. In TDI flutter remote sensing imaging, due to the push-broom characteristic of TDI cameras, the blur kernels of each row of the flutter image differ. Using row-by-row restoration can significantly improve the image restoration effect. When the difference in blur kernels between rows is small, appropriately increasing the number of image blocks to be restored in the current row is beneficial to improving the image restoration effect; however, when the difference in blur kernels between rows is large, increasing the number of image block rows has no significant effect on the image restoration effect. Increasing the number of image block rows will prolong the overall image restoration time. By constructing a similarity formula between blur kernels, the degree of difference in blur kernels between rows of the flutter image is compared to determine the size of the image block required to restore the current row.

[0004] This patent analyzes the impact of image patch length and the similarity value of the inter-line blur kernel on the restoration of TDI remote sensing flutter images, constructs a formula relating image patch length to the similarity value of the inter-line blur kernel, and proposes a method to adaptively adjust the image patch length based on the similarity value of the inter-line blur kernel, thereby optimizing the restoration effect. The focus of this article is on the restoration of image flutter, without addressing the geometric detection and analysis of image flutter.

[0005] The paper "Image Dither Reduction Algorithm Based on Histogram Matching" (Journal of Shanxi Datong University, Vol. 33, No. 2, 2017) discloses an image dither reduction algorithm. It designs a color histogram matching algorithm based on pre-defined target matching regions (one or more). By analyzing and comparing the similarity of the color histograms of the target and candidate regions in two images before and after dithering, the optimal matching block is found, and the corresponding block's motion displacement is obtained. This allows for the calculation of the absolute motion displacement of the two images, ultimately eliminating the interference displacement. However, the method described in this paper requires a standard image before dithering as a reference, making it unsuitable for dithering analysis of single-frame images and limiting its application scenarios. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for detecting and correcting flutter in satellite remote sensing images.

[0007] According to the present invention, a method and system for detecting and correcting satellite remote sensing image jitter are provided, the scheme of which is as follows:

[0008] In a first aspect, a method for detecting and correcting flutter in satellite remote sensing images is provided, the method comprising:

[0009] Step S1: Select ground features as the analysis objects in the remote sensing image;

[0010] Step S2: Perform edge detection and extraction on the analyzed object;

[0011] Step S3: Fit the edge detection results;

[0012] Step S4: Identify the flutter error of the remote sensing image based on the fitting results;

[0013] Step S5: Perform frequency analysis on the flutter error of the remote sensing image to identify the frequency characteristics of the flutter error;

[0014] Step S6: Perform image correction based on remote sensing image resampling.

[0015] Preferably, the analysis object selected in step S1 is a land feature with obvious straight-line characteristics in the remote sensing image.

[0016] Preferably, in step S2, edge detection is performed by locating the feature regions of light and dark transitions in the image and identifying the edge lines of the light and dark transitions through polynomial interpolation.

[0017] Preferably, the fitting method for the edge detection results in step S3 is polynomial fitting, which is adapted to ground features with curve features.

[0018] Preferably, the method for identifying flutter error in remote sensing images in step S4 is to obtain the wave characteristics of ground objects in the remote sensing image by subtracting the fitting result from the detection result, and use this as the image flutter error.

[0019] Secondly, a satellite remote sensing image flutter detection and correction system is provided, the system comprising:

[0020] Module M1: Select ground features as the analysis object in remote sensing images;

[0021] Module M2: Performs edge detection and extraction on the analyzed object;

[0022] Module M3: Fits the results of edge detection;

[0023] Module M4: Identifies the flutter error of the remote sensing image based on the fitting results;

[0024] Module M5: Performs frequency analysis on the flutter error of remote sensing images and identifies the frequency characteristics of the flutter error;

[0025] Module M6: Image correction based on remote sensing image resampling method.

[0026] Preferably, the analysis object selected in module M1 is a land feature with obvious straight-line characteristics in the remote sensing image.

[0027] Preferably, the edge detection in module M2 is achieved by locating the feature regions of the image's light and dark transitions and identifying the edge lines of the light and dark transitions through polynomial interpolation.

[0028] Preferably, the fitting method for the edge detection results in module M3 adopts polynomial fitting, thereby adapting to ground features with curve features.

[0029] Preferably, the flutter error identification method for remote sensing images in module M4 is to obtain the wave characteristics of ground objects in the remote sensing image by subtracting the fitting result and the detection result, and use this as the image flutter error.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. This invention can identify objects using image information without relying on external measurement data, making it easy to promote and apply.

[0032] 2. This invention detects and corrects flutter in satellite remote sensing images, which is of great significance for improving the quality of remote sensing images. Attached Figure Description

[0033] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0034] Figure 1 This is a flowchart of high-resolution image flutter detection and correction.

[0035] Figure 2 This is a schematic diagram of image flutter;

[0036] Figure 3 It is the result of image edge contour extraction;

[0037] Figure 4 This is the result of image flutter correction. Detailed Implementation

[0038] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0039] This invention provides a method for detecting and correcting flutter in satellite remote sensing images, referring to... Figure 1 As shown, the method specifically includes:

[0040] Step S1: Select ground features as the analysis objects in the remote sensing image;

[0041] Step S2: Perform edge detection and extraction on the analyzed object;

[0042] Step S3: Fit the edge detection results;

[0043] Step S4: Identify the flutter error of the remote sensing image based on the fitting results;

[0044] Step S5: Perform frequency analysis on the flutter error of the remote sensing image to identify the frequency characteristics of the flutter error;

[0045] Step S6: Perform image correction based on remote sensing image resampling.

[0046] In step S1, the selected analysis object is a land feature with obvious straight-line characteristics in the remote sensing image, such as an overpass or an airport runway.

[0047] In step S2, the edge detection method identifies the edge lines of the light-dark transition by locating the feature regions of the image's light-dark transition through polynomial interpolation.

[0048] In step S3, the fitting method for the edge detection results adopts polynomial fitting, which can adapt to ground features with curve features and improve the flutter detection accuracy.

[0049] In step S4, the flutter error identification method for remote sensing images is to obtain the wave characteristics of ground objects in the remote sensing image by subtracting the fitting result from the detection result, which is then used as the image flutter error.

[0050] The present invention will now be described in more detail.

[0051] A method for detecting and correcting flutter in satellite remote sensing images, referring to Figure 1 and Figure 2 As shown, the details are as follows:

[0052] Step S1: Select ground features as the analysis objects in the remote sensing image;

[0053] This invention does not rely on external measurement information; instead, it utilizes internal image information to identify image flutter. The principle is to identify the contour lines of reference objects within the image and detect the fluctuations in these contour lines as the image flutter value. Therefore, when selecting the analysis object, features with obvious straight-line characteristics should be chosen, such as overpasses and airport runways.

[0054] Step S2: Perform edge detection and extraction on the analysis object; the principle of edge contour extraction is to locate the area of ​​light and dark variation of the analysis object and extract it through polynomial interpolation or centroid method. The specific steps are as follows:

[0055] A. Obtain the pixel coordinates and grayscale values ​​of areas with varying brightness in the image. For example, continuously obtain the grayscale values ​​of n pixels, and denote the grayscale value of the i-th pixel as DN. i ;

[0056] B. Identify the pixel with the largest gray value and the pixel with the smallest gray value in the sequence.

[0057] The pixel with the largest grayscale value is denoted as m, and the grayscale value is DN. m The pixel with the smallest grayscale value is denoted as n, and its grayscale value is DN. n ;

[0058] C. Calculate the average of the largest and smallest grayscale pixels. The formula is as follows:

[0059]

[0060] Then, search for the three pixels in the sequence whose gray values ​​are closest to the average gray value, and label them k1, k2, and k3, with their corresponding gray values ​​DN. k1 DN k2 and DN k3 .

[0061] D. Perform a second-order linear fit on the three pixels numbered k1, k2, and k3. The fitting function is denoted as:

[0062] y = a·x 2 +b·x+c

[0063] In the above fitting function, x is the gray value corresponding to the pixel;

[0064] y is the number corresponding to the pixel;

[0065] a, b, and c are the fitting coefficients.

[0066] E. Calculate the pixel number of the edge contour. The calculation formula is as follows:

[0067] y k =a·DNmean 2 +b·DN mean +c

[0068] In the above formula, y k The cell number is used to represent the edge contour.

[0069] Following the above process, by processing the pixel sequence of each line column in the image, the edge contour coordinates of the ground features can be obtained, as shown in the attached figure. Figure 3 As shown.

[0070] Step S3: Fit the edge detection results;

[0071] The contour coordinates identified in step S2 are fitted. Considering that the actual ground features have a certain degree of curvature, a third-order polynomial is chosen for fitting, and the fitting function is denoted as:

[0072] y1=p1·x1 3 +p2·x1 2 +p3·x1+p4

[0073] In the above formula, x1 is the x-direction pixel coordinate of the edge contour, y1 is the y-direction pixel coordinate of the edge contour, and p1, p2, p3 and p4 are fitting coefficients.

[0074] Step S4: Identify the flutter error of the remote sensing image based on the fitting results;

[0075] By subtracting the measured edge contours from the curve fitting results, the wave characteristics of the ground feature contours are identified and used as image flutter error. The specific calculation expression is as follows:

[0076] Erro i =p1·x i 3 +px·x i 2 +p3·x i +p4-y i

[0077] In the above formula, Erro i Let be the image flutter error of the i-th pixel of the ground feature;

[0078] x i Let be the measured x-direction pixel coordinates of the i-th pixel of the ground feature;

[0079] y i Let be the measured x-direction pixel coordinates of the i-th pixel of the ground feature;

[0080] p1, p2, p3, and p4 are the contour curve fitting coefficients.

[0081] Step S5: Perform Fourier analysis on the flutter error of the remote sensing image identified in Step S4 to identify the frequency of the flutter in the remote sensing image, and compare it with the vibration source on the satellite to identify the vibration source on the satellite, providing a basis for the future optimization design of the vibration isolation structure of the satellite.

[0082] Step S6: Perform image correction based on remote sensing image resampling.

[0083] Reference Figure 4 As shown, image correction is performed based on the image jitter error identified above. In order to achieve high-precision correction at the sub-pixel level, pixel resampling is used to obtain the corrected image and eliminate the influence of image jitter.

[0084] This invention provides a method for detecting and correcting jitter in satellite remote sensing images. This invention can identify jitter without relying on external measurement data, and can be completed through image information, which is convenient for widespread application. By detecting and correcting jitter in satellite remote sensing images, it is of great significance for improving the quality of remote sensing images.

[0085] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0086] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for detecting and correcting flutter in satellite remote sensing images, characterized in that, include: Step S1: Select ground features as the analysis objects in the remote sensing image; Step S2: Perform edge detection and extraction on the analyzed object; In step S2, edge detection involves locating feature regions where light and dark transitions occur in the image and identifying the edge lines of these transitions using polynomial interpolation. The specific steps are as follows: A. Obtain the pixel coordinates and grayscale values ​​of the areas of varying brightness in the image, and continuously acquire the values ​​for each line column. The gray value of the i-th pixel is denoted as _i_th ...i_i_ The grayscale value of each pixel is ; B. Identify the pixel with the largest gray value and the pixel with the smallest gray value in the sequence; The pixel with the largest gray value is denoted as grayscale value The pixel with the smallest grayscale value is denoted as . grayscale value ; C. Calculate the average of the largest and smallest grayscale pixels. The formula is as follows: Then, search for the three pixels in the sequence whose gray values ​​are closest to the average gray value, and number them as follows: and The corresponding grayscale value is denoted as , and ; D. For the numbered and A second-order linear fit is performed on the three pixels, and the fitting function is denoted as: In the above fitting function, This represents the grayscale value corresponding to the pixel. The number corresponding to the pixel; , and These are the fitting coefficients; E. Calculate the pixel number of the edge contour. The calculation formula is as follows: In the above formula, The cell number for the edge contour; By processing the pixel sequence of each line column in the image according to the above process, the edge contour coordinates of the ground features can be obtained. Step S3: Fit the edge detection results; Step S3 further includes: fitting the contour coordinates identified in step S2; considering that the actual ground features have a certain degree of curvature, a third-order polynomial is selected for fitting, and the fitting function is denoted as: In the above formula, For the edge contour Orientation cell coordinates For the edge contour Orientation cell coordinates , , and These are the fitting coefficients; Step S4: Identify the flutter error of the remote sensing image based on the fitting results; Step S4 involves subtracting the measured edge contour from the curve fitting result to identify the wave characteristics of the ground feature contour, which is used as the image flutter error; the specific calculation expression is as follows: In the above formula, For the first land feature Image flutter error per pixel; For the first land feature Actual measurement of each pixel Orientation cell coordinates; For the first land feature Actual measurement of each pixel Orientation cell coordinates; , , and These are the fitting coefficients for the contour curve; Step S5: Perform frequency analysis on the flutter error of the remote sensing image to identify the frequency characteristics of the flutter error; Step S6: Perform image correction based on remote sensing image resampling.

2. The satellite remote sensing image jitter detection and correction method according to claim 1, characterized in that, The analysis object selected in step S1 is a land feature with obvious straight-line characteristics in the remote sensing image.

3. A satellite remote sensing image flutter detection and correction system, characterized in that, include: Module M1: Select ground features as the analysis object in remote sensing images; Module M2: Performs edge detection and extraction on the analyzed object; In module M2, edge detection involves locating feature regions where light and dark transitions occur in the image and identifying the edge lines of these transitions using polynomial interpolation. The specific steps are as follows: A. Obtain the pixel coordinates and grayscale values ​​of the areas of varying brightness in the image, and continuously acquire the values ​​for each line column. The gray value of the i-th pixel is denoted as _i_th ...i_i_ The grayscale value of each pixel is ; B. Identify the pixel with the largest gray value and the pixel with the smallest gray value in the sequence; The pixel with the largest gray value is denoted as grayscale value The pixel with the smallest grayscale value is denoted as . grayscale value ; C. Calculate the average of the largest and smallest grayscale pixels. The formula is as follows: Then, search for the three pixels in the sequence whose gray values ​​are closest to the average gray value, and number them as follows: and The corresponding grayscale value is denoted as , and ; D. For the numbered and A second-order linear fit is performed on the three pixels, and the fitting function is denoted as: In the above fitting function, This represents the grayscale value corresponding to the pixel. The number corresponding to the pixel; , and These are the fitting coefficients; E. Calculate the pixel number of the edge contour. The calculation formula is as follows: In the above formula, The cell number for the edge contour; By processing the pixel sequence of each line column in the image according to the above process, the edge contour coordinates of the ground features can be obtained. Module M3: Fits the results of edge detection; The module M3 further includes: fitting the contour coordinates identified by module M2; considering that the actual ground features have a certain degree of curvature, a third-order polynomial is selected for fitting, and the fitting function is denoted as: In the above formula, For the edge contour Orientation cell coordinates For the edge contour Orientation cell coordinates , , and These are the fitting coefficients; Module M4: Identifies the flutter error of the remote sensing image based on the fitting results; The module M4 identifies the fluctuation characteristics of the ground feature contour by subtracting the measured edge contour results from the curve fitting results, which is used as the image flutter error; the specific calculation expression is as follows: In the above formula, For the first land feature Image flutter error per pixel; For the first land feature Actual measurement of each pixel Orientation cell coordinates; For the first land feature Actual measurement of each pixel Orientation cell coordinates; , , and These are the fitting coefficients for the contour curve; Module M5: Performs frequency analysis on the flutter error of remote sensing images and identifies the frequency characteristics of the flutter error; Module M6: Image correction based on remote sensing image resampling method.

4. The satellite remote sensing image flutter detection and correction system according to claim 3, characterized in that, The analysis object selected in module M1 is a land feature with obvious straight-line characteristics in a remote sensing image.

Images