A method for evaluating the degree of degradation caused by atmospheric factors on optical system imaging

Abstract

The present invention relates to the field of image degradation processing technology, and specifically to a method for evaluating the degree of degradation caused by atmospheric factors to optical system imaging. The present invention constructs a systematic evaluation system for the impact of atmospheric factors on optical imaging. By integrating the modeling process of foggy aerosol MTF and long-exposure turbulence MTF, using the Nyquist frequency as a bridge, the atmospheric coherence length when the two are equal is solved, achieving a breakthrough from independent factor analysis to quantification of synergistic effects. An evaluation model is constructed by integrating multiple parameters such as optical thickness, atmospheric coherence length, wavelength, and spatial angular frequency. Combined with Fourier transform image simulation and evaluation indicators such as peak signal-to-noise ratio and structural similarity index, a complete technical chain of "modeling-simulation-evaluation" is formed, providing a solution that combines theoretical rigor and engineering practicality for predicting the degree of degradation of optical imaging systems in complex atmospheric environments.

Description

Technical Field

[0001] The present invention relates to the technical field of image degradation processing, and in particular to a method for evaluating the degree of degradation caused by atmospheric factors on optical system imaging. Background Art

[0002] In high-end optical imaging fields such as space remote sensing, astronomical observation, and military reconnaissance, the interaction between atmospheric media and light waves is a key bottleneck limiting image quality. Aerosol scattering (such as fog and dust) and turbulent disturbances (such as random fluctuations in the atmospheric refractive index) are the two core physical factors that degrade image quality. Aerosols cause broadband light attenuation through the Mie scattering mechanism, resulting in overall image distortion such as reduced contrast and blurred edges. Turbulence, on the other hand, distorts image detail structure by distorting the light wavefront phase, causing dynamic flicker and resolution loss. The modulation transfer function (MTF), an authoritative metric for quantifying the frequency domain response of an imaging system, can effectively separate the two effects. Aerosol MTF exhibits a low-frequency-dominated attenuation characteristic, while long-exposure turbulence MTF exhibits a high-frequency exponential decay trend. After decades of theoretical development and technological accumulation, the academic community has established independent and comprehensive research frameworks around aerosol MTF and long-exposure turbulence MTF. Aerosol research is based on the small-angle scattering approximation model and, through classical theories such as Koschmieder's law, establishes the relationship between visibility and optical depth. τ The quantitative relationship between atmospheric coherence length and atmospheric coherence length is widely used in atmospheric environment monitoring and weather forecasting; turbulence research relies on the Fried long exposure model and uses professional equipment such as the Differential Image Motion Monitor (DIMM) to achieve the atmospheric coherence length. ρ The precise measurement of 0 provides important parameter support for ground-based astronomical observations.

[0003] However, aerosols and turbulence in the natural atmospheric environment do not exist in isolation, and the complex synergistic effect of the two significantly exacerbates the resolution loss of the optical system. For example, dust particles, as special aerosols, have an uneven spatial distribution, which will cause changes in local turbulence intensity; and turbulent motion accelerates the diffusion and aggregation of aerosol particles, forming a dynamic coupling effect. The current research method calculates the effects of aerosols and turbulence independently, separating the interaction mechanism between the two, resulting in the inability to accurately predict image degradation in actual scenes, that is, it is impossible to more accurately determine the sensitivity of atmospheric factors to optical imaging. This shows that constructing a theoretical model that can accurately describe the synergistic effect of aerosols and turbulence has become a key scientific issue in breaking through the bottleneck of high-end optical imaging technology. Summary of the Invention

[0004] To avoid and overcome the technical problems of existing technologies, the present invention provides a method for assessing the degree of degradation caused by atmospheric factors on optical system imaging. This method leverages the interaction between foggy aerosols and long-exposure turbulence to accurately predict image degradation in real-world scenarios.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors includes the following:

[0007] 1. MTF

[0008] 1. MTF of aerosol in foggy weather

[0009] In order to accurately quantify the degradation effect of scattering and absorption of foggy sol on imaging quality, the optical thickness, wavelength, spatial angular frequency and light transmission distance of the optical system are considered, and a physical model of foggy sol MTF is established based on the small-angle scattering approximation theory, as shown in formula (1):

[0010] (1);

[0011] (2);

[0012] (3);

[0013] (4);

[0014] (5);

[0015] (6);

[0016] Where, Indicates foggy aerosol MTF, represents the spatial angular frequency; Expressed as a natural constant e An exponential function with base ; represents pi; Indicates the optical transmission distance, Indicates the maximum optical transmission distance; represents the scattering angle, Indicates the maximum scattering angle; Indicates wavelength; represents the scattering coefficient function; represents the Henyey-Greenstein scattering phase function; represents an asymmetric factor function; represents the Bessel function; represents the differential of the scattering angle; Represents the differential of light transmission distance; represents the extinction coefficient; represents the single scattering albedo; Indicates visibility; represents the visibility coefficient; represents the optical thickness; represents the asymmetry factor in the scattering phase function.

[0017] 2. Long Exposure Turbulence MTF

[0018] In order to accurately describe the image degradation mechanism caused by long-exposure turbulence, the wavelength, spatial angular frequency and atmospheric coherence length of the optical system are considered, and the long-exposure turbulence MTF obtained by Fried using the light propagation theory in random media is adopted, as shown in formula (7):

[0019] (7);

[0020] (8);

[0021] Where, represents the long exposure turbulence MTF; represents the atmospheric coherence length; represents the wave number; Represents the long-exposure turbulent refractive index structure constant Path distribution; represents the long exposure turbulence transmission distance; Indicates distance The differential of .

[0022] 2. Atmospheric MTF Equivalent Image Degradation Model

[0023] The Nyquist frequency of the optical system was obtained, and the foggy aerosol MTF and long-exposure turbulence MTF values ​​of the optical system were calculated. The corresponding Nyquist frequency, optical depth, wavelength, and atmospheric coherence length (ACL) were obtained when the foggy aerosol MTF and long-exposure turbulence MTF values ​​of the optical system were equal. A calculation model for the ACL was constructed using the Nyquist frequency, optical depth, and wavelength as independent variables and the ACL as the dependent variable. An equivalent model was established using the ratio of the dimensionless parameter to the optical depth. The ACL calculation model and the equivalent model were combined to form an atmospheric MTF equivalent image degradation model.

[0024] 1. The construction process of the atmospheric coherence length calculation model is as follows:

[0025] S21. Obtain the focal length and pixel size of the optical system and calculate the value of the Nyquist frequency , the calculation formula is as follows:

[0026] (9);

[0027] Where, F Indicates focal length, in mm; P Indicates the pixel size in mm;

[0028] S22, calculated is the spatial angular frequency, which is substituted into the foggy aerosol MTF and long-exposure turbulence MTF to form an equivalent equation:

[0029] ;

[0030] S23, determine the optical thickness, wavelength, maximum light transmission distance, maximum scattering angle and scattering phase function parameters of the optical system, together with Substitute the equivalent equation, solve the corresponding atmospheric coherence length, and form a set of training samples ;

[0031] S24. According to the contents of steps S21 to S23, training samples of different optical systems are obtained, and the Nyquist frequency, optical thickness and wavelength are used as independent variables, and the atmospheric coherence length is used as the dependent variable. A calculation model of the atmospheric coherence length is constructed by power function fitting, which is specifically expressed as follows:

[0032] (10);

[0033] Where, It indicates the atmospheric coherence length of the optical system when the aerosol MTF value in foggy weather and the turbulence MTF value in long exposure are equal.

[0034] 2. The construction process of the equivalent model is as follows:

[0035] According to the sampling theorem, the effective resolution of the optical system is limited by the Nyquist frequency. Through theoretical analysis, it can be seen that the effects of foggy aerosol scattering and long-exposure turbulence on images show obvious frequency band characteristics. and In terms of:

[0036] (1) Below the Nyquist frequency, aerosol scattering in foggy weather is the dominant factor causing image degradation.

[0037] (2) Above the Nyquist frequency, the phase disturbance caused by long exposure turbulence is the main reason for the decrease in system resolution.

[0038] Based on this physical phenomenon, the present invention proposes a dimensionless parameter , that is, at the Nyquist frequency, the foggy aerosol corresponding to a specific optical thickness is equivalent to The corresponding long exposure turbulence MTF has an effect on the image degradation. Theoretical analysis shows that when the wavelength is constant, for different optical systems, the optical thickness is the same, and the dimensionless parameter This characteristic provides an important theoretical basis for establishing an equivalent atmospheric MTF image degradation model. As the dependent variable, an equivalent model is established by exponential fitting, which is specifically expressed as follows:

[0039] (11);

[0040] 3. Image Simulation

[0041] The foggy weather aerosol MTF value of the optical system is calculated using the foggy weather aerosol MTF value, and the original image of the optical system is simulated using the foggy weather aerosol MTF value to obtain the foggy weather aerosol simulation image. The image distortion of the foggy weather aerosol simulation image is evaluated by the peak signal-to-noise ratio. The atmospheric coherence length of the optical system is calculated using the atmospheric coherence length calculation model, and the long-exposure turbulence MTF value of the optical system is calculated in combination with the long-exposure turbulence MTF. The original image of the optical system is simulated using the long-exposure turbulence MTF value to obtain the long-exposure turbulence simulation image. The image quality of the long-exposure turbulence simulation image is evaluated by the structural similarity index.

[0042] 1. The evaluation process of the image distortion of the foggy aerosol simulation image is as follows:

[0043] S3A1. Determine the wavelength, maximum light transmission distance, maximum scattering angle, scattering phase function parameters, focal length, and pixel size of the optical system, calculate the Nyquist frequency of the optical system, and calculate the foggy aerosol MTF value when the spatial angular frequency is the Nyquist frequency. At the same time, obtain the original image of the optical system.

[0044] S3A2, based on the MTF value of foggy weather aerosol, and combined with Fourier transform, the original image is converted into a foggy weather aerosol simulation image.

[0045] S3A3. Obtain the grayscale value of each pixel in the original image and the foggy weather sol simulation image, and use the grayscale value to calculate the peak signal-to-noise ratio for evaluating the image distortion of the foggy weather sol simulation image. The peak signal-to-noise ratio is calculated as follows:

[0046] (12);

[0047] (13);

[0048] Where, represents the peak signal-to-noise ratio; represents the logarithmic function with base 10; Represents the maximum grayscale value in the original image; Represents the grayscale mean between the original image and the foggy weather sol simulation image; Indicates the total number of pixel rows in the original image; Indicates the total number of pixel columns of the original image; Indicates the original image Row and Grayscale value at the intersection of the columns; Indicates the first Row and Grayscale value at the intersection of the columns; express and The square of the 2nd-order norm between .

[0049] 2. The image quality evaluation process of the long exposure turbulence simulation image is as follows:

[0050] S3B1. Based on the optical system in step S3A1, a corresponding atmospheric coherence length is calculated using an atmospheric coherence length calculation model.

[0051] S3B2. Use the atmospheric coherence length to calculate the long-exposure turbulence MTF value when the spatial angular frequency is the Nyquist frequency.

[0052] S3B3. Based on the long-exposure turbulence MTF value and combined with Fourier transform, the original image is converted into a long-exposure turbulence simulation image.

[0053] S3B4. Obtain the grayscale value of each pixel in the original image and the long-exposure turbulence simulation image, and use the grayscale value to calculate a structural similarity index for evaluating the image quality of the long-exposure turbulence simulation image. The calculation formula of the structural similarity index is as follows:

[0054] (14);

[0055] Where, x represents the original graph; y Represents a long exposure turbulence simulation image; Represents the structural similarity index between the original image and the long-exposure turbulence simulation image; Represents the grayscale mean in the original image; Represents the grayscale mean in the long exposure turbulence simulation image; Represents the grayscale standard deviation in the original image; Indicates the grayscale standard deviation in the long exposure turbulence simulation image; express and The covariance between and All represent constants to avoid the denominator being 0, among which, , , and are 0.01 and 0.03 respectively, Indicates the dynamic range of the image.

[0056] 4. Image Degradation Assessment

[0057] An image degradation assessment model is established based on image distortion and image quality, and the image degradation degree caused by foggy aerosol and long-exposure turbulence on the optical system is evaluated by the image degradation assessment model.

[0058] The process of establishing the image degradation assessment model is as follows:

[0059] S41. Based on the principle of linear analysis, the following prediction equation is constructed:

[0060] (15);

[0061] Where, Indicates the degree of image degradation; represents the image distortion weight coefficient; Represents the image quality weight coefficient.

[0062] S42. Setting the values ​​of two weight coefficients in the prediction equation to convert the prediction equation into an image degradation assessment model.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] 1. This invention establishes a systematic evaluation system for the impact of atmospheric factors on optical imaging. By integrating the modeling processes of foggy aerosol MTF and long-exposure turbulence MTF, and using the Nyquist frequency as a bridge to solve the atmospheric coherence length when the two are equivalent, this system achieves a breakthrough from independent factor analysis to quantification of synergistic effects. Its innovation lies in integrating multiple parameters such as optical depth, atmospheric coherence length, wavelength, and spatial angular frequency to construct an evaluation model. Combining Fourier transform image simulation with evaluation metrics such as peak signal-to-noise ratio and structural similarity index, this system forms a complete "modeling-simulation-evaluation" technology chain. This system provides a solution that combines theoretical rigor with engineering practicality for predicting the degree of degradation of optical imaging systems in complex atmospheric environments.

[0065] 2. MTF of foggy aerosol is based on the Henyey-Greenstein scattering phase function and Mie scattering theory, and builds a refined broadband light attenuation model by integrating the scattering angle and light transmission distance. Its advantage lies in the introduction of visibility piecewise function to dynamically adjust the wavelength dependence of the extinction coefficient, and achieve accurate modeling for different foggy transparency; through the scattering phase function parameters g By describing the forward scattering characteristics of foggy aerosol particles and combining the coupled calculation of optical thickness and transmission distance, the attenuation mechanism of foggy aerosol on the low-frequency components of the image is carefully depicted, providing an accurate mathematical tool for visibility compensation in remote sensing imaging.

[0066] 3. The long-exposure turbulence MTF is based on Fried's long-exposure theory, with the atmospheric coherence length as its core parameter. It accurately characterizes the suppression of high-frequency signals by the optical wavefront phase distortion caused by long-exposure turbulence through the exponential decay of wavelength and spatial angular frequency to the 5 / 3 power of the coherence length. Its innovation lies in calculating the coherence length through the path integral of the wavenumber and the refractive index structure constant of the long-exposure turbulence. This model, which considers the dynamic changes in long-exposure turbulence intensity at different altitudes, is suitable for the performance evaluation of ground-based high-resolution imaging systems and can accurately reflect the image detail distortion and resolution loss caused by long-exposure turbulence.

[0067] 4. The atmospheric coherence length calculation model constructs a prediction model that combines physical significance with data-driven characteristics by fitting multiple training samples with a power function. This breakthrough lies in solving the critical conditions for the equalization of image degradation between the aerosol MTF and the long-exposure turbulence MTF in foggy weather, quantifying the dominance boundary of these two factors in different imaging systems, and enabling rapid calculation of the atmospheric coherence length under synergistic effects. Compared to traditional integration methods, this model significantly improves computational efficiency and maintains low prediction error over a wide frequency range, making it suitable for real-time assessment in multiple scenarios, including both spaceborne and ground-based applications.

[0068] 5. The frequency-domain attenuation characteristics of foggy aerosol MTF are mapped to the spatial domain through Fourier transform, generating a simulated image containing low-frequency attenuation features. The peak signal-to-noise ratio (PSNR) is then used to quantify the distortion. This method's advantage lies in accurately capturing the image contrast loss and overall blurring caused by foggy aerosols through the logarithmic transformation of the mean square error of grayscale values ​​and the maximum grayscale value. Furthermore, by combining the effect of wavelength on the extinction coefficient, the attenuation effect can be independently assessed for different wavelengths (e.g., visible light and near-infrared), providing a quantifiable evaluation standard for remote sensing image restoration and multispectral imaging in foggy and hazy weather.

[0069] 6. The Structural Similarity Index (SSIM) is introduced to evaluate the quality of long-exposure turbulence simulation images. By combining grayscale mean, standard deviation, and covariance, it accurately captures image detail distortion and structural damage caused by long-exposure turbulence phase distortion. Compared to traditional pixel-level error metrics, it effectively reflects the degree of loss of image structural information, providing a scientific measure of the correction effect of adaptive optics systems.

[0070] 7. Using linearly weighted PSNR and SSIM metrics, a comprehensive evaluation system was constructed that integrates foggy aerosol distortion and long-exposure turbulent structural damage. This innovation lies in the dynamic adjustment of weighting coefficients based on the imaging scenario (e.g., spaceborne remote sensing prioritizes global clarity, while military reconnaissance emphasizes detail recognition). This allows for precise mapping of degradation levels, directly guiding the optimization of imaging system parameters (e.g., band selection and sampling frequency adjustment), and providing a quantitative basis for determining the operating mode of optical systems under different atmospheric conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 Flowchart of the present invention.

[0072] Figure 2 In the present invention, the Nyquist frequency is 5000 rad. -1 The equilibrium point curves of long-exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0073] Figure 3 In the present invention, the Nyquist frequency is 10000rad -1 The equilibrium point curves of long exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0074] Figure 4 In the present invention, the Nyquist frequency is 25000rad -1 The equilibrium point curves of long exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0075] Figure 5 In the present invention, the Nyquist frequency is 50000rad -1 The equilibrium point curves of long exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0076] Figure 6 In the present invention, the Nyquist frequency is 100000rad -1 The equilibrium point curves of long exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0077] Figure 7 In the present invention, the Nyquist frequency is 250000rad -1The equilibrium point curves of long exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0078] Figure 8 In the present invention, the Nyquist frequency is 500000rad -1 The equilibrium point curves of long exposure turbulence MTF and foggy aerosol MTF are shown in Figure 2.

[0079] Figure 9 For the present invention and the Nyquist frequency and The power function fitting curve of .

[0080] Figure 10 For the present invention and Exponential fitting curve graph.

[0081] Figure 11 This is the original black and white target grid image in the present invention.

[0082] Figure 12 This is the first set of simulation diagrams in the present invention.

[0083] Figure 13 This is the second set of simulation diagrams in the present invention. DETAILED DESCRIPTION

[0084] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0085] When the target is transmitted to the imaging system image plane through the atmosphere (including long-exposure turbulence, foggy aerosol), the random motion of the long-exposure turbulence will cause irregular changes in the atmospheric refractive index, and the scattering and absorption of foggy aerosol will cause the quality of the beam reaching the image plane to attenuate, thereby causing effects such as wavefront distortion and image blur, ultimately leading to image quality degradation. Figure 1 As shown, the present invention proposes a method for evaluating the degree of degradation caused by atmospheric factors on optical system imaging, which specifically includes the following contents.

[0086] 1. Establishing MTF

[0087] 1. MTF of aerosol in foggy weather

[0088] In order to accurately quantify the degradation effect of scattering and absorption of foggy aerosol on imaging quality, a physical model of foggy aerosol MTF is established based on the small-angle scattering approximation theory, as shown in formula (1).

[0089] 2. Long Exposure Turbulence MTF

[0090] In order to accurately describe the image degradation mechanism caused by long-exposure turbulence, the present invention adopts the long-exposure turbulence MTF obtained by Fried using the light propagation theory in random media, as shown in formula (7).

[0091] 2. Establishing an Atmospheric MTF Equivalent Image Degradation Model

[0092] When the optical systems have the same Nyquist frequency, the quantitative effects of foggy aerosol MTF and long-exposure turbulence MTF on image degradation remain constant, even if the focal length and pixel size vary. This application selected focal lengths of 100 mm, 500 mm, and 1000 mm, and pixel sizes of 1 mm, 50 mm, and 100 mm for image analysis, and used Equation (9) to calculate the Nyquist frequency. The specific parameters and calculation results are detailed in Table 1.

[0093] Table 1 Optical system parameters and Nyquist frequency under classical conditions

[0094] ;

[0095] According to formula (1) and formula (7), the corresponding optical thickness and atmospheric coherence length are calculated when the foggy aerosol MTF values ​​at the seven Nyquist frequencies listed in Table 1 are equal to the long-exposure turbulence MTF values. The calculation results are as follows: Figures 2 to 8 shown.

[0096] When the long exposure turbulence MTF is equal to the foggy aerosol MTF at the Nyquist frequency (i.e. Figures 2 to 8 The corresponding curve intersection point in the As the critical atmospheric coherence length , which characterizes the equilibrium state between the effects of long-exposure turbulence and foggy aerosol on the imaging performance of optical systems. This characteristic provides a quantitative basis and threshold reference for evaluating the dominant factors of different degradation effects in atmospheric channels.

[0097] The power function fitting is used to obtain and 、 , the atmospheric coherence length calculation model of Nyquist frequency, such as Figure 9 As shown, to calculate a specific Nyquist frequency, and under . Figure 9 The fit curve in the function represents the fitting curve.

[0098] Before the Nyquist frequency, the foggy aerosol MTF decays rapidly to a stable value. The corresponding long exposure turbulence MTF decays rapidly in the middle and high frequencies, so The influence of the long exposure turbulence MTF corresponding to the divisor of on the image is equivalent to the foggy aerosol MTF. Combining the data in Table 2 and using exponential fitting to establish divisor of and An equivalent model such as Figure 10 shown. Figure 10 The fit curve in represents the fitted curve, and original data represents the original data. The above atmospheric coherence length calculation model and the equivalent model are combined to form the atmospheric MTF equivalent image degradation model.

[0099] Table 2 Degradation conditions of equivalent images of foggy aerosol and long-exposure turbulence at Nyquist frequency

[0100] ;

[0101] Depend on Figures 2 to 8 It can be seen that:

[0102] (1) As the Nyquist frequency increases, the specific optical thickness Corresponding It gradually increases, indicating that the imaging performance of the optical system is more sensitive to long-exposure turbulence.

[0103] (2) The MTF of foggy aerosol decays rapidly to a fixed value at low spatial angular frequencies, and its contribution to image degradation is weakly affected by system parameters.

[0104] Optical thickness corresponding to the MTF of foggy aerosol at the Nyquist frequency Corresponding to the long exposure turbulence MTF There are significant regular features between them.

[0105] Depend on Figures 2 to 8 It can be seen that when When , the Nyquist frequency data set is as follows:

[0106] Nyquist = [5000,10000,25000,50000,100000,250000,500000];

[0107] ;

[0108] ;

[0109] ;

[0110] ;

[0111] ;

[0112] ;

[0113] By fitting the above data set with a power function, we obtain the following analytical expression:

[0114] (16);

[0115] Figure 9 Shows the relationship between the Nyquist frequency, The power function fitting curve is shown in the figure. According to the calculation, the mean square error of the model is 1.82×10 -3 , the average relative error is 3.59%.

[0116] Based on the model established by formula (16), the expressions of the atmospheric coherence length calculation model at different wavelengths are derived, as shown in formula (10).

[0117] According to formula (10), the cross-characteristics of the foggy aerosol MTF corresponding to a specific optical thickness and the long-exposure turbulence MTF corresponding to different atmospheric coherence lengths at different Nyquist frequencies and wavelengths can be quantitatively analyzed to determine the attenuation magnitude relationship between the two at the Nyquist frequency, thereby establishing the corresponding atmospheric coherence length calculation model.

[0118] In order to verify the accuracy of the atmospheric MTF equivalent image degradation model, that is, to evaluate the image degradation degree of foggy aerosol and long exposure turbulence is consistent and The magnitude of the two different Nyquist frequencies (5000 rad -1 、500000 rad -1 ) optical system is used for image simulation. The distortion and image quality of the simulation results are quantitatively evaluated by the peak signal-to-noise ratio (PSNR) shown in formula (12) and the structural similarity index (SSIM) shown in formula (14).

[0119] 3. Image Simulation

[0120] In order to intuitively characterize the degradation of image quality caused by foggy aerosol and long-exposure turbulence, a black and white target grid image (image size: 110×220 pixels) with both initial contrast and modulation contrast of 1 is constructed. The maximum surface reflectivity of the grid image is 1 and the minimum is 0, as shown in the following example: Figure 11 shown.

[0121] This section simulates the different foggy conditions under the conditions of 550nm wavelength. Corresponding foggy aerosol MTF and The corresponding long exposure turbulence MTF pair Figure 11 degree of degradation.

[0122] First, the foggy aerosol MTF is calculated based on formula (1), where: Take 1.0, Take 0.8 corresponding to foggy days, They are 3.52, 1.76, and 0.77 respectively. Image simulations are performed for optical systems with different parameters.

[0123] Optical system 1: focal length 10 cm, pixel size 10 mm, Nyquist frequency 5000 rad -1 .

[0124] Optical system 2: focal length 100 cm, pixel 1 mm, Nyquist frequency 500,000 rad -1 .

[0125] Then, according to formula (10), the corresponding values ​​of the two optical systems are calculated respectively. :

[0126] The calculated results for Group 1 are: 0.1280cm, 0.1831cm, and 0.3018cm.

[0127] The calculated results for Group 2 are: 13.3635cm, 20.2553cm, and 33.3868cm.

[0128] Combined with formula (11), we get is 7.5, 6.0, and 4.0, and the atmospheric MTF equivalent image degradation model can be obtained. :

[0129] The first group was 0.0171cm, 0.0305cm, and 0.0755cm; the second group was 1.7818cm, 3.3759cm, and 8.3467cm.

[0130] Finally, the long exposure turbulence MTF is calculated according to formula (7), and the foggy aerosol MTF value and the image MTF of the original image are simulated based on the above calculation. The simulation results are as follows: Figure 12 and Figure 13 shown.

[0131] Figure 12The figure shows the MTF imaging simulation results of the first group of black and white target grid images. The first column shows the foggy weather aerosol simulation image, the second column shows the long-exposure turbulence simulation image, and the third column shows the simulation image under the synergistic effect of foggy weather aerosol and long-exposure turbulence. The data under each simulation image is PSNR / SSIM.

[0132] Figure 13 The figure shows the MTF imaging simulation results of the second group of black and white target grid images. The first column shows the foggy weather aerosol simulation image, the second column shows the long-exposure turbulence simulation image, and the third column shows the simulation image under the synergistic effect of foggy weather aerosol and long-exposure turbulence. The data under each simulation image is PSNR / SSIM.

[0133] Depend on Figure 12 、 Figure 13 It can be seen that:

[0134] (1) With From 3.52 to 0.77, the image degradation degree gradually decreases, the stripe clarity improves, and the PSNR and SSIM indicators improve accordingly; at the same time, the long exposure turbulence that meets the same degradation effect An increasing trend indicates When is reduced, the long-exposure turbulence intensity required to achieve equivalent image degradation is relatively weak.

[0135] (2) The foggy sol MTF is not sensitive to the change of Nyquist frequency. The image degradation effects of the optical systems of Group 1 and Group 2 under the same foggy sol conditions are basically the same. However, the long-exposure turbulence MTF is highly sensitive to the change of Nyquist frequency. When the Nyquist frequency of the optical system increases, the image degradation effect of the optical system of Group 1 and Group 2 under the same foggy sol conditions is basically the same. The corresponding conditions becomes larger, and the turbulence intensity becomes weaker with long exposure.

[0136] (3) When When the turbulence intensity is less than 1cm, the long exposure is extremely strong. Figure 12 It can be seen that the Nyquist frequency is 5000rad -1 When the Nyquist frequency increases to 500,000 rad, the effect of long exposure turbulence on the image is weaker than that of foggy sol. -1 When the exposure time is long, the impact of turbulence on the image is significantly enhanced.

[0137] 4. Image Degradation Assessment

[0138] Based on formula (15), the influence of foggy aerosol and long exposure turbulence on the image is described, where the weight coefficient is and The calculation results are shown in Table 3.

[0139] Table 3 Error analysis of optical system simulation result index P

[0140] ;

[0141] The accuracy of the model was verified by using two image quality indicators, PSNR and SSIM, and the influence of long-exposure turbulence and foggy aerosol on image degradation under different optical system parameters was analyzed.

[0142] As shown in Table 3, the absolute error of the proposed image degradation assessment model under different optical system conditions is less than 0.2, and the relative error does not exceed 3%. This shows that the image degradation assessment model has good adaptability and versatility under different optical system parameters.

[0143] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors, characterized in that: The following steps are involved: S1. Establish foggy aerosol MTF based on the optical thickness, wavelength, spatial angular frequency and light transmission distance of the optical system; establish long-exposure turbulence MTF based on the wavelength, spatial angular frequency and atmospheric coherence length of the optical system; S2. Obtain the Nyquist frequency of the optical system and calculate the foggy weather aerosol MTF value and long-exposure turbulence MTF value of the optical system; obtain the corresponding Nyquist frequency, optical thickness, wavelength, and atmospheric coherence length when the foggy weather aerosol MTF value and the long-exposure turbulence MTF value of the optical system are equal; and construct an atmospheric coherence length calculation model using the Nyquist frequency, optical thickness, and wavelength as independent variables and the atmospheric coherence length as the dependent variable; S3. Calculate the foggy weather aerosol MTF value of the optical system using the foggy weather aerosol MTF value, simulate the original image of the optical system using the foggy weather aerosol MTF value to obtain a foggy weather aerosol simulation image, and evaluate the image distortion of the aerosol simulation image using the peak signal-to-noise ratio; calculate the atmospheric coherence length of the optical system using the atmospheric coherence length calculation model, calculate the long-exposure turbulence MTF value of the optical system in combination with the long-exposure turbulence MTF, simulate the original image of the optical system using the long-exposure turbulence MTF value to obtain a long-exposure turbulence simulation image, and evaluate the image quality of the long-exposure turbulence simulation image using the structural similarity index; S4. An image degradation assessment model is established based on image distortion and image quality, and the image degradation degree caused by foggy aerosol and long-exposure turbulence on the optical system is evaluated by the image degradation assessment model.

2. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 1, characterized in that: The MTF of foggy aerosol is expressed as follows: ； Where, Indicates foggy aerosol MTF, represents the spatial angular frequency; Expressed as a natural constant e An exponential function with base ; represents pi; Indicates the optical transmission distance, Indicates the maximum optical transmission distance; represents the scattering angle, Indicates the maximum scattering angle; Indicates wavelength; represents the scattering coefficient function; represents the Henyey-Greenstein scattering phase function; represents an asymmetric factor function; represents the Bessel function; represents the differential of the scattering angle; Represents the differential of light transmission distance; ； ； ； ； ； Where, represents the extinction coefficient; represents the single scattering albedo; Indicates visibility; represents the visibility coefficient; represents the optical thickness; represents the asymmetry factor in the scattering phase function.

3. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 2, characterized in that: The long exposure turbulence MTF is expressed as follows: ； ； Where, represents the long exposure turbulence MTF; represents the atmospheric coherence length; represents the wave number; Represents the long-exposure turbulent refractive index structure constant Path distribution; represents the long exposure turbulence transmission distance; Indicates distance The differential of .

4. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 3, characterized in that: The construction process of the atmospheric coherence length calculation model is as follows: S21. Obtain the focal length and pixel size of the optical system and calculate the value of the Nyquist frequency , the calculation formula is as follows: ； Where, F Indicates focal length, in mm; P Indicates the pixel size in mm; S22, calculated is the spatial angular frequency, which is substituted into the foggy aerosol MTF and long-exposure turbulence MTF to form an equivalent equation: ； S23, determine the optical thickness, wavelength, maximum light transmission distance, maximum scattering angle and scattering phase function parameters of the optical system, together with Substitute the equivalent equation, solve the corresponding atmospheric coherence length, and form a set of training samples ; S24. According to the contents of steps S21 to S23, training samples of different optical systems are obtained, and the atmospheric coherence length calculation model is constructed by power function fitting with Nyquist frequency, optical thickness and wavelength as independent variables and atmospheric coherence length as dependent variable.

5. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 4, characterized in that: The atmospheric coherence length calculation model is expressed as follows: ； Where, It indicates the atmospheric coherence length of the optical system when the aerosol MTF value in foggy weather and the turbulence MTF value in long exposure are equal.

6. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 5, characterized in that: The evaluation process of the image distortion of the foggy aerosol simulation image is as follows: S3A1. Determine the wavelength, maximum light transmission distance, maximum scattering angle, scattering phase function parameters, focal length, and pixel size of the optical system, calculate the Nyquist frequency of the optical system, and calculate the foggy weather aerosol MTF value when the spatial angular frequency is the Nyquist frequency, and simultaneously obtain the original image of the optical system; S3A2, based on the foggy weather sol MTF value, combined with Fourier transform, the original image is converted into a foggy weather sol simulation image; S3A3. Obtain the grayscale value of each pixel in the original image and the foggy weather sol simulation image, and use the grayscale value to calculate the peak signal-to-noise ratio for evaluating the image distortion of the foggy weather sol simulation image.

7. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 6, characterized in that: The peak signal-to-noise ratio is calculated as follows: ； ； Where, represents the peak signal-to-noise ratio; represents the logarithmic function with base 10; Represents the maximum grayscale value in the original image; Represents the grayscale mean between the original image and the foggy weather sol simulation image; Indicates the total number of pixel rows in the original image; Indicates the total number of pixel columns of the original image; Indicates the original image Row and Grayscale value at the intersection of the columns; Indicates the first Row and Grayscale value at the intersection of the columns; express and The square of the 2nd-order norm between .

8. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 7, characterized in that: The evaluation process of the long-exposure turbulence simulation image quality is as follows: S3B1, based on the optical system in step S3A1, calculate the corresponding atmospheric coherence length using an atmospheric coherence length calculation model; S3B2. Calculate the long-exposure turbulence MTF value when the spatial angular frequency is the Nyquist frequency using the atmospheric coherence length. S3B3, based on the long-exposure turbulence MTF value, combined with Fourier transform, the original image is converted into a long-exposure turbulence simulation image; S3B4. Obtain the grayscale value of each pixel in the original image and the long-exposure turbulence simulation image, and use the grayscale value to calculate a structural similarity index for evaluating the image quality of the long-exposure turbulence simulation image.

9. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 8, characterized in that: The calculation formula of the structural similarity index is as follows: ； Where, x represents the original graph; y Represents a long exposure turbulence simulation image; Represents the structural similarity index between the original image and the long-exposure turbulence simulation image; Represents the grayscale mean in the original image; Represents the grayscale mean in the long exposure turbulence simulation image; Represents the grayscale standard deviation in the original image; Indicates the grayscale standard deviation in the long exposure turbulence simulation image; express and The covariance between and All represent constants.

10. The method for evaluating the degree of degradation of optical system imaging caused by atmospheric factors according to claim 9, characterized in that: The process of establishing the image degradation assessment model is as follows: S41. Based on the principle of linear analysis, the following prediction equation is constructed: ； Where, Indicates the degree of image degradation; represents the image distortion weight coefficient; Represents the image quality weight coefficient; S42. Setting the values ​​of two weight coefficients in the prediction equation to convert the prediction equation into an image degradation assessment model.

Images