A method and device for detecting the dynamic modulation transfer function of a lunar exploration camera on orbit

Abstract

The present invention belongs to the field of image processing technology, and specifically relates to a method and device for detecting the dynamic modulation transfer function of a lunar exploration camera on orbit. The method first obtains the observation image of the lunar exploration camera based on the observation area mission planning, and determines whether the observation image contains a target area and the type of the target area contained therein; wherein the types of target areas include edge areas and star point areas, the edge areas contain edge targets, and the high-contrast edge formed by the edge of the moon and the deep sky background area or the high-contrast edge formed by the shadow area and the illuminated area on the moon serves as the edge target, and the star point area contains luminous star targets; if the type of the target area is an edge area, the edge method is used to obtain the MTF; if the type of the target area is a star point area, the star point method is used to obtain the MTF. The present invention has the advantages of low cost, high precision, multiple measurements, good real-time performance, and high precision of detection results.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical system imaging quality evaluation and processing, and specifically relates to an on-orbit dynamic modulation transfer function detection method and device for a lunar exploration camera. Background Art

[0002] During the on-orbit operation of the lunar exploration camera, it is affected by factors such as the space environment (optical system environmental adaptability, camera reliability, thermal control environment, TDICCD driving capability, attitude changes and yaw angle drift, etc.) and the aging of its own components. These factors will lead to a decline in image quality, resulting in a decrease in the amount of information that the image can provide. Therefore, it is necessary to conduct regular on-orbit inspections of the lunar exploration camera's imaging quality to monitor its performance changes. How to dynamically evaluate the imaging quality of the lunar exploration camera during its on-orbit operation is a major problem that needs to be solved. The modulation transfer function (MTF) is an important parameter for evaluating the imaging quality of the optical system. It is a direct, objective, accurate and quantitative camera image quality evaluation indicator. It can reflect the degree of information attenuation during the imaging of the object by the optical system to be evaluated. It has high accuracy and is more objective and comprehensive than traditional image quality evaluation methods. It is an ideal evaluation indicator for the on-orbit performance of the lunar exploration camera. Currently, the MTF test of lunar exploration cameras is mainly carried out in the laboratory before the launch of the probe. The MTF is calculated using a modulation transfer function meter or by calculating the camera's MTF through image information obtained by imaging some specific targets. However, there will be a certain error between the MTF calculated in the laboratory and the actual MTF of the lunar exploration camera during its in-orbit operation. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and device for detecting the dynamic modulation transfer function of a lunar exploration camera on orbit, so as to solve the technical problem of poor detection accuracy of the MTF detection results when performing MTF detection on lunar exploration cameras in the prior art.

[0004] To solve the above technical problems, the present invention provides an on-orbit dynamic MTF detection method for a lunar exploration camera, comprising the following steps:

[0005] 1) Based on the observation area mission planning, the observation image of the lunar exploration camera is acquired, and the observation image is judged to determine whether the target area is included in the observation image and the type of the target area included. The target area types include edge areas and star areas. The edge areas contain edge targets, and the high-contrast edge formed by the edge of the moon and the deep sky background area or the high-contrast edge formed by the shadow area and the illuminated area on the moon is regarded as the edge target. The star area contains luminous star targets.

[0006] 2) If the target area is a knife-edge area, the knife-edge method is used to obtain the MTF; if the target area is a star point area, the star point method is used to obtain the MTF.

[0007] The beneficial effects of the above technical solution are as follows: the on-orbit dynamic MTF detection method for the lunar exploration camera of the present invention rationally utilizes the observation target and space target characteristics of the lunar exploration camera itself, uses the high-contrast edge formed by the edge of the moon and the deep-space background area or the high-contrast edge formed by the shadow area and the illuminated area on the moon as the edge target, and uses the deep-space star point-shaped target as the star point target. There is no need to design and install special detection instruments, nor is there any need to lay lunar surface targets. It can conveniently realize real-time on-orbit detection of MTF, effectively saving manpower, material resources, and financial resources, and facilitating direct, objective, accurate, and quantitative evaluation of camera image quality. Moreover, the on-orbit dynamic MTF detection method for the lunar exploration camera of the present invention has the advantages of low cost, high precision, and multiple measurements. The method of the present invention does not calculate the MTF of the lunar exploration camera in the laboratory, but detects its MTF when the lunar exploration camera is in orbit, with good real-time performance and high precision of the detection results.

[0008] Furthermore, the star point method is used to obtain the modulation transfer function by: performing brightness peak detection on multiple star observation images to detect the sub-pixel star point center position; performing encrypted registration of the star point light spot data grayscale values ​​based on the star point center positions detected in multiple images to obtain encrypted star sampling points; fitting the encrypted star sampling points to obtain the point spread function (PSF); obtaining the image's MTF discrete points with respect to frequency based on the obtained PSF and fitting the MTF discrete points to obtain the MTF.

[0009] The beneficial effect of the above technical solution is that the grayscale values ​​of the star point spots are encrypted and registered according to the center positions of the star points detected in multiple images, which can greatly improve the PSF fitting accuracy.

[0010] Furthermore, the means of obtaining the MTF discrete points of the image with respect to frequency is as follows: discretize x within a certain width interval and set the y value equal to 0 to obtain a number of discrete points (x, y), and then calculate the discrete value of the PSF (x, y) of each discrete point (x, y), where x represents the distance from each pixel point to the star point in the image, and y represents the image grayscale value of each pixel point in the image; based on the discrete value of the PSF (x, y), perform a one-dimensional Fourier transform within the width interval, and then take the modulus and normalize it to obtain the MTF discrete points of the image with respect to frequency.

[0011] Furthermore, a two-dimensional Gaussian surface function is used to fit the encrypted star sampling points.

[0012] Furthermore, a Gaussian function is used to fit the MTF discrete points.

[0013] Furthermore, the knife-edge method is used to obtain the MTF by: performing edge detection on the target area and fitting the knife-edge curve; after fitting the knife-edge curve, calculating the vertical distance x from all pixels in the image to the knife-edge curve point by point; fitting the knife-edge data [x, y] to obtain the edge spread function ESF of the target area, where y represents the grayscale value of each pixel in the image; taking the derivative of the obtained ESF to obtain the line spread function LSF; obtaining the MTF discrete points of the image with respect to frequency based on the obtained LSF and fitting the MTF discrete points to obtain the MTF.

[0014] Furthermore, the method for obtaining the discrete points of the MTF of the image with respect to frequency is as follows: discretizing x within a certain width interval, and obtaining the corresponding discrete values ​​of the line spread function LSF based on the discrete x; performing a one-dimensional Fourier transform on the obtained discrete values ​​of the line spread function LSF, and then taking the modulus and normalizing them to obtain the discrete points of the MTF of the image with respect to frequency.

[0015] Furthermore, the method further includes calculating the MTF value at the Nyquist frequency.

[0016] Furthermore, the observation area mission is obtained based on at least one of the lunar probe's orbit prediction data, lunar topography reference data, star catalog data, and lunar ephemeris data.

[0017] Its beneficial effects are: planning the observation mission based on the lunar probe's orbit prediction data, lunar topography reference data, star catalog data and lunar ephemeris data, so as to select the best time and target area for observing the moon.

[0018] In order to solve the above technical problems, the present invention also provides an on-orbit dynamic modulation transfer function detection device for a lunar exploration camera, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the on-orbit dynamic modulation transfer function detection method for a lunar exploration camera of the present invention is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flow chart of the method steps of the present invention;

[0020] Figure 2 It is a full-color band lunar image acquired by the observation satellite;

[0021] Figures 3 to 6 These are different examples of lunar crater and ridge images;

[0022] Figure 7 This is a schematic diagram of the principle of calculating MTF using the knife-edge method of the present invention;

[0023] Figure 8 is a schematic diagram of the blade edge image selected in the up and down directions;

[0024] Figure 9 is a schematic diagram of the blade edge image selected in the left and right directions;

[0025] Figure 10 It is aimed at Figure 8 Schematic diagram of ESF curve;

[0026] Figure 11 It is aimed at Figure 8 Schematic diagram of LSF curve;

[0027] Figure 12 It is aimed at Figure 8 Schematic diagram of the MTF curve;

[0028] Figure 13 It is aimed at Figure 9 Schematic diagram of ESF curve;

[0029] Figure 14 It is aimed at Figure 9 Schematic diagram of LSF curve;

[0030] Figure 15 It is aimed at Figure 9 Schematic diagram of the MTF curve;

[0031] Figure 16 It is a schematic diagram of the principle of calculating MTF using the point source method of the present invention. DETAILED DESCRIPTION

[0032] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0033] An embodiment of an on-orbit dynamic modulation transfer function detection method for a lunar exploration camera:

[0034] In this embodiment, MTF refers to modulation transfer function; ESF refers to edge spread function; LSF refers to line spread function; and PSF refers to point spread function.

[0035] like Figure 1 As shown, the on-orbit dynamic modulation transfer function detection method of the lunar exploration camera of the present invention includes the following steps:

[0036] Step 1: Use the high-contrast edge formed by the edge of the moon and the deep-sky background area, or the high-contrast edge formed by the shadow area and the illuminated area on the moon as the knife-edge target, and use the deep-sky star-shaped target as the star point target; use the knife-edge target and the star point target as the target area, and select the area containing more target areas on the global image of the moon as the target observation object.

[0037] The edge of the moon and the deep sky background area (the moon edge is bright, the deep sky background is dark) form a high contrast edge, such as the edge of the target. Figure 2 As shown, among them, Figures 3 to 6 As shown, shadowed areas on the Moon consist of lunar impact craters or ridges.

[0038] Step 2: Plan the observation mission based on the orbit prediction data of the lunar probe, lunar topography reference data, star catalog data, lunar ephemeris data and other information, and select the best time and target area for observation to obtain observation images.

[0039] The observation mission is planned based on the lunar probe's orbit prediction data, lunar topography reference data, star catalog data, and lunar ephemeris data, making it easier to select the best time and target area for lunar observation.

[0040] Step 3: Determine whether the observed image contains the target area. If not, the MTF detection ends. If the target area is contained, further determine the type of the target area. If the target area is a knife-edge target, the knife-edge method is used to obtain the MTF. If the target area is a star point target, the star point method is used to obtain the MTF.

[0041] The circular edge target formed by the edge of the moon and the deep space background is an ideal natural edge target. Among the earth observation satellites, the Gaofen-4 satellite, the Fengyun-2 satellite, and the Fengyun-4 satellite have all conducted MTF on-orbit testing based on lunar observations, with good results. During the lunar exploration process, the edge of the moon can also be imaged to obtain edge images that meet the MTF calculation requirements and perform on-orbit MTF testing. Figure 7 As shown in FIG, the method for obtaining MTF using the knife-edge method includes the following steps:

[0042] 1) Perform edge detection on the target area and fit the edge curve. That is, select an image, detect the edge points of the target area line by line, and fit the edge points to obtain the edge curve.

[0043] 2) After fitting the knife edge curve, the vertical distance x from each pixel in the image to the knife edge curve is calculated point by point.

[0044] 3) Fit the edge data [x, y] to obtain the edge spread function (ESF) of the target area, where y is the grayscale value of each pixel in the image. The least squares method can be used for fitting. The expression for fitting the edge spread function when using the least squares method is as follows:

[0045]

[0046] Where a i 、b i 、ci , D is a parameter to be determined, and ESF(x) represents the function value of the edge spread function at x.

[0047] 4) Derivative the edge spread function to obtain the line spread function LSF. The expression of the obtained line spread function LSF is as follows:

[0048]

[0049] Where a i 、b i 、c i , D are parameters to be determined.

[0050] 5) Obtain the image's MTF discrete points with respect to frequency based on the obtained line spread function LSF and fit the MTF discrete points to obtain the MTF. The specific process of obtaining the MTF discrete points is as follows:

[0051] ① Discretize x according to the first preset width x value interval and the first preset step size, and obtain the corresponding discrete value of the line spread function based on the discretized x. Here, x∈[-30,30] can be selected as the first preset width x value interval, and the first preset step size can be selected as 0.05.

[0052] ② Perform a one-dimensional Fourier transform on the discrete values ​​of the line spread function, then modulo and normalize them to obtain the MTF discrete points of the image with respect to frequency. The expression of the obtained MTF discrete points is as follows:

[0053]

[0054] Where, MTF(n) _Normalize represents the nth MTF discrete point, where n is the discrete point number; FFT represents the fast Fourier transform function.

[0055] ③ When fitting the MTF discrete points, a Gaussian function can be used for fitting. The fitted MTF expression is as follows:

[0056]

[0057] Where a, b, and c are the parameters to be fitted, f is the frequency, and MTF(f) represents the MTF function value at point f.

[0058] 6) Calculate the Nyquist frequency f _Nyquist The MTF value at Nyquist frequency is half of the cutoff frequency, and the calculation formula is as follows:

[0059]

[0060] When selecting an image in step 1), you can select it from the top and bottom directions or from the left and right directions. The image selected from the top and bottom directions is as follows: Figure 8 As shown, the images selected from the left and right directions are as follows Figure 9 If the image is selected from the top and bottom directions, the ESF curve obtained in step 3) is as follows Figure 10 As shown, the LSF curve obtained in step 4) is as follows Figure 11 As shown, the MTF curve obtained in step 5) is as follows Figure 12 As shown; if the image is selected from the left and right directions, the ESF curve obtained in step 3) is as follows Figure 13 As shown, the LSF curve obtained in step 4) is as follows Figure 14 As shown, the MTF curve obtained in step 5) is as follows Figure 15 shown.

[0061] The MTF detection method based on star points takes the luminous stars under the dark background of space as point source targets. During the lunar exploration process, according to the predicted orbit of the lunar probe, combined with the star catalog, and taking into account the avoidance of the sun, the information of stars appearing in the instrument's observable field of view (including the star magnitude, the time and position of appearance in the instrument coordinate system) is predicted, and then multiple star observations are performed to obtain star point images and perform MTF calculations. The light intensity distribution of the star point can be expressed by the point spread function PSF. The point spread function is symmetrical about the central circle. Usually, the brightness distribution along the x-axis is used as the point spread function LSF. The modulus of its Fourier transform result is the modulation transfer function MTF. Its principle is as follows Figure 16 As shown in the figure, the PSF obtained by using the point source response of a single image usually has insufficient sampling points and is easily affected by factors such as noise, resulting in large errors in the obtained PSF. This algorithm uses a method of collecting multiple star point images at a time and aligning and encrypting the multiple images to calculate the MTF. The algorithm steps are as follows:

[0062] 1) Perform brightness peak detection on multiple star observation images to obtain the sub-pixel star point center position.

[0063] 2) Based on the center positions of the star points detected in multiple images, the grayscale values ​​of the star point spot data are encrypted and registered to obtain encrypted star sampling points. This encrypted registration of the grayscale values ​​of the star point spot data is beneficial to improving the point spread function (PSF) fitting accuracy.

[0064] 3) Use the two-dimensional Gaussian surface function to fit the encrypted star sampling points to obtain a high-precision point spread function (PSF). The expression of the obtained point spread function is:

[0065]

[0066] Where σ xis the standard deviation in the x direction, σ y is the standard deviation in the y direction, μ x is the mean value in the x direction, μ y is the mean value in the y direction, x is the distance from each pixel in the image to the star point, y is the image grayscale value of each pixel in the image, and PSF(x,y) is the PSF function value at the point (x,y).

[0067] 4) Obtain the image's MTF discrete points with respect to frequency based on the obtained point spread function and fit the MTF discrete points to obtain the MTF. When fitting the MTF discrete points, a Gaussian function can be used for fitting, and its expression is as follows:

[0068]

[0069] Where a, b, and c are the parameters to be fitted, f is the frequency, and MTF(f) represents the MTF function value at frequency f.

[0070] The specific process of obtaining the MTF discrete points about frequency is as follows:

[0071] ① Take an x-value interval of the second preset width and a second preset step size, discretize x, set the y value to 0, obtain a number of discrete points (x, y), and calculate the discrete values ​​of PSF(x, y) based on the discrete points (x, y). The x-value interval of the second preset width can be selected from x∈[-30, 30], and the second preset step size can be selected from 0.05.

[0072] ② Perform a one-dimensional Fourier transform on the discrete values ​​of the line spread function, then modulate and normalize them to obtain the MTF discrete points of the image with respect to frequency.

[0073] 5) Calculate the Nyquist frequency f _Nyquist The MTF value at Nyquist frequency is half of the cutoff frequency, and the calculation formula is as follows:

[0074]

[0075] The on-orbit dynamic MTF detection method for a lunar exploration camera of the present invention rationally utilizes the observation target and space target characteristics of the lunar exploration camera itself, uses the high-contrast edge formed by the edge of the moon and the deep-space background area, or the high-contrast edge formed by the shadow area and the illuminated area on the moon, as the edge target, and uses deep-space star-shaped targets as the star point targets. It does not require the design and installation of special detection instruments, nor does it require the laying of lunar surface targets. It can conveniently achieve real-time on-orbit MTF detection, effectively saving manpower, material resources, and financial resources. The on-orbit dynamic MTF detection method for a lunar exploration camera of the present invention has the advantages of low cost, high accuracy, and the ability to perform multiple measurements. The method of the present invention does not calculate the MTF of the lunar exploration camera in a laboratory, but rather detects the MTF of the lunar exploration camera while it is in orbit, resulting in good real-time performance and high accuracy in detection results.

[0076] An embodiment of an on-orbit dynamic modulation transfer function detection device for a lunar exploration camera:

[0077] The present invention provides an on-orbit dynamic modulation transfer function detection device for a lunar exploration camera, comprising a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the on-orbit dynamic modulation transfer function detection method for a lunar exploration camera of the present invention is implemented. The specific details and technical effects of the on-orbit dynamic MTF detection method for a lunar exploration camera of the present invention have been described in detail in the embodiment of the on-orbit dynamic MTF detection method for a lunar exploration camera and will not be repeated here. The processor may be a processing device such as a microprocessor MCU or a programmable logic device FPGA; the memory may be any type of memory that uses electrical energy to store information.

Claims

1. A method for detecting the dynamic modulation transfer function of a lunar exploration camera on orbit, characterized in that: include: 1) Based on the observation area mission plan, obtain observation images from the lunar exploration camera and determine whether the observation images contain target areas and the types of target areas included. Target areas include edge areas and star areas. Knife-edge areas contain edge targets, and high-contrast edges formed by the edge of the moon and the deep sky background, or high-contrast edges formed by shadow areas and illuminated areas on the moon, are considered edge targets. Star areas contain luminous star targets. 2) If the target area is a knife-edge area, the following knife-edge method is used to obtain the MTF: brightness peak detection is performed on multiple star observation images to detect the sub-pixel star center position; the grayscale value of the star spot data is encrypted and registered based on the detected star center position to obtain encrypted star sampling points; the encrypted star sampling points are fitted to obtain the point spread function (PSF); based on the PSF, the image's MTF discrete points with respect to frequency are obtained and the MTF discrete points are fitted to obtain the MTF; If the target area is a star point area, the following star point method is used to obtain the MTF: edge detection is performed on the target area, and a knife edge curve is fitted; then the vertical distance from each pixel in the image to the knife edge curve is calculated. x ; Edge data[ x , y ] is fitted to obtain the edge spread function ESF of the target area, y is the grayscale value of each pixel in the image; the line spread function LSF is obtained by derivative of ESF; the MTF discrete points of the image with respect to frequency are obtained based on LSF and the MTF discrete points are fitted to obtain MTF.

2. The on-orbit dynamic modulation transfer function detection method of the lunar exploration camera according to claim 1 is characterized in that: The method of obtaining the MTF discrete points of the image with respect to frequency is: x Discretize and let y The value is equal to 0, and several discrete points are obtained ( x , y ), and then calculate each discrete point ( x , y )of PSF( x , y ), where x represents the distance from each pixel in the image to the star point, and y represents the image grayscale value of each pixel in the image; according to the PSF ( x , y ), perform one-dimensional Fourier transform in the width interval, then modulo and normalize it to obtain the MTF discrete points of the image with respect to frequency.

3. The on-orbit dynamic modulation transfer function detection method of the lunar exploration camera according to claim 1, characterized in that: The encrypted star sampling points are fitted using a two-dimensional Gaussian surface function.

4. The on-orbit dynamic modulation transfer function detection method of the lunar exploration camera according to claim 1, characterized in that: A Gaussian function is used to fit the MTF discrete points.

5. The on-orbit dynamic modulation transfer function detection method of the lunar exploration camera according to claim 1, characterized in that: The method of obtaining the MTF discrete points of the image with respect to frequency is: x Discretize, based on the discrete x Obtain the corresponding discrete value of the line spread function LSF; perform a one-dimensional Fourier transform on the obtained discrete value of the line spread function LSF, then take the modulus and normalize it to obtain the MTF discrete points of the image with respect to frequency.

6. The on-orbit dynamic modulation transfer function detection method of the lunar exploration camera according to claim 1, characterized in that: It also calculates the MTF value at the Nyquist frequency.

7. The on-orbit dynamic modulation transfer function detection method of the lunar exploration camera according to claim 1, characterized in that: The observation area mission is obtained based on at least one of the orbit prediction data of the lunar probe, lunar topography reference data, star catalog data and lunar ephemeris data.

8. An on-orbit dynamic modulation transfer function detection device for a lunar exploration camera, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that: The computer program is used to implement the on-orbit dynamic modulation transfer function detection method for a lunar exploration camera as described in any one of claims 1 to 7 when executed by a processor.

Images