A wide swath SAR image intensity equalization method based on adaptive parameter adjustment

Abstract

The present application relates to a kind of wide SAR image intensity equalization method based on adaptive parameter adjustment, by analyzing radar imaging power equation, establish the corresponding target scattering intensity variation curve in range dimension, and on this basis, combined with azimuth pattern characteristics, fit out azimuth intensity variation curve, and combine it with the scattering intensity variation curve in range, utilize the changed curve to carry out two-dimensional adjustment to image intensity, i.e. wide SAR image of brightness variation equalization can be obtained.The present application method realizes the image intensity equalization of wide SAR image, and then effectively improves image quality, and lays a good foundation for subsequent SAR image interpretation and detection identification.

Description

Technical Field

[0001] This invention belongs to the field of imaging technology, specifically relating to a wide-swath SAR image intensity equalization method based on adaptive parameter adjustment. Background Technology

[0002] Synthetic Aperture Radar (SAR) imaging technology achieves high azimuth resolution by creating a virtual large aperture through platform movement, thereby obtaining high-resolution remote sensing images of the observed scene. Since the brightness variations of scattering points in SAR images are determined by the scattering characteristics of the observed target, image quality is influenced not only by the imaging algorithm but also by electromagnetic scattering characteristics. This is especially true when radar performs wide-area observations; due to the large image coverage, the intensity of scattering points at distant points decreases significantly, leading to a reduction in the accuracy of the observed target's scattering characteristics. Therefore, after acquiring SAR images, it is necessary to perform image scattering intensity equalization based on system characteristics and electromagnetic scattering features to improve image readability.

[0003] Considering that the brightness variation of observed targets in radar images is mainly reflected by the target's scattering intensity, the image brightness is primarily determined by the radar's power and the electromagnetic scattering characteristics of the observation scene. Given that the main influencing factors for range and azimuth intensity variations differ, separate analysis and processing are required for each dimension. In the range direction, the intensity of image scattering points is mainly affected by the intensity of the radar's emitted electromagnetic waves and its own scattering characteristics. Therefore, we consider combining traditional radar equations, deriving and transforming them, and re-establishing a range-direction image intensity variation curve suitable for radar imaging, thereby achieving balanced intensity adjustment in the range direction. For the azimuth direction, its swath width is mainly determined by the radar's azimuth beamwidth, and its scattering intensity is mainly related to the corresponding azimuth antenna pattern gain. Therefore, we establish a corresponding image azimuth intensity variation curve by fitting and adjusting the system pattern gain curve. Based on this, image intensity is adjusted, thereby effectively improving image brightness balance and enhancing image quality.

[0004] Traditional synthetic aperture radar (SAR) utilizes the principle of electromagnetic scattering to acquire the scattering intensity variation characteristics of a target scene, thereby obtaining SAR image results of the observed scene. Based on electromagnetic scattering characteristics, when the observed scene area is small, the echo intensity within the radar beam coverage area can be considered to accurately reflect the scattering characteristics of target points within the scene. That is, the brightness variations of scattering points in the image are determined by the scattering characteristics of the observed target, resulting in images with high readability and realism. However, when conducting wide-area observations, due to the attenuation characteristics of electromagnetic waves, the energy of the illumination echo at distant points in the image decreases significantly. This not only causes a decrease in brightness at distant points but also prevents the accurate reflection of the scattering characteristics of the observed target, hindering subsequent interpretation and analysis and reducing the intelligence value of SAR. Therefore, after imaging, it is necessary to re-equalize the overall scattering intensity of the image based on system performance and electromagnetic scattering characteristics to improve the image's realism and readability.

[0005] Since the brightness variation of SAR images is mainly reflected by the intensity of scattering points in the observed scene, in the range direction, the intensity of scattering points is affected not only by their own scattering characteristics but also by the intensity of the electromagnetic waves emitted by the radar at that point. Traditional radar equations establish an analytical expression for the relationship between radar signal-to-noise ratio (SNR) and range. Because the scattering intensity of SAR images is essentially equivalent to the SNR, we consider rewriting the radar equations to establish a range-direction image intensity variation curve, thereby achieving a balanced range-direction image intensity variation. In the azimuth direction, considering that the azimuth swath width is usually determined by the radar beamwidth, its scattering intensity variation is mainly determined by the azimuth antenna pattern gain curve. Therefore, we consider fitting and adjusting the pattern gain curve to obtain the azimuth image intensity variation curve, achieving balanced azimuth intensity adjustment and effectively improving the overall intensity variation and brightness uniformity of the SAR image. Summary of the Invention

[0006] Technical problems to be solved

[0007] To address the issue of gradually decreasing image brightness caused by beamwidth limitations and electromagnetic scattering characteristics during wide-area observation and imaging using airborne synthetic aperture radar, this invention provides a wide-swath SAR image intensity equalization method based on adaptive parameter adjustment.

[0008] Technical solution

[0009] A wide-swath SAR image intensity equalization method based on adaptive parameter adjustment, characterized by the following steps:

[0010] Step 1: Establish the range-based scattering intensity variation curve h r (R img );

[0011]

[0012] P a =P t ·ratio 占

[0013] ratio 占 =τ / PRT

[0014]

[0015] Among them, P t Let A be the peak transmitted power of the radar, A be the antenna gain area, and η be the antenna gain area. A The antenna efficiency is represented by λ, the radar transmission wavelength is represented by v, the aircraft speed is represented by K and T0, which represent the Boltzmann constant and standard room temperature, respectively. n L is the system noise figure. s With C B Representing the system loss and bandwidth correction factor, respectively, α is the viewing angle under the radar imaging beam, and R c N is the pixel distance from the center point of the image. r ρ represents the number of pixels in the distance dimension of the SAR image. r σ is the range resolution of SAR imaging. ρ R is the scattering coefficient. img This represents the range of distance variation in the image.

[0016] Step 2: Based on the azimuth pattern characteristics, fit the azimuth scattering intensity variation curve h. a (N a );

[0017] 2a) Pre-acquire the azimuth pattern curve corresponding to the imaging mode of the radar system;

[0018] 2b) Calculate the corresponding oblique angle using the inertial navigation velocity parameters and system parameters;

[0019]

[0020] Where PRF is the system pulse repetition frequency, and asin(·) represents the operation of taking a sine angle;

[0021] 2c) Determine the required angle range θ of the direction pattern to be captured based on the oblique viewing angle. pattern ∈[-θ s ,θ s Extract the gain variation curve of the antenna pattern within the corresponding angular range;

[0022] 2d) The gain variation curve is interpolated and fitted to the same length as the number of azimuth points in the image to obtain the azimuth scattering intensity variation curve h. a (N a ), where N aThis represents the total number of points in the direction of orientation;

[0023] Step 3: Establish a two-dimensional compensation curve to balance the image brightness;

[0024] 3a) Take the absolute value of the original image data to obtain the corresponding grayscale image I(R). img N a ),

[0025] 3b) Multiply the range scattering intensity variation curve with the azimuth scattering intensity variation curve to obtain the two-dimensional image scattering intensity equalization matrix S. eq (R img N a );

[0026] 3c) Equalize the scattering intensity matrix S of the two-dimensional image eq (R img N a Compensate to the original grayscale image I(R) img N a In this process, the two are multiplied together to obtain the final intensity-equalized grayscale image.

[0027] A further technical solution of the present invention: the interpolation method in step 2 is sinc interpolation or linear interpolation.

[0028] A further technical solution of the present invention: Step 3b) further includes analyzing the range scattering intensity variation curve h. r (R img Preprocessing: Invert the curve, then normalize it, and finally take the cube root of the normalized curve.

[0029] A further technical solution of the present invention: Step 3b) also includes the azimuth scattering intensity variation curve h. a (N a Preprocessing: Invert the curve, then normalize it, and finally window or square root the normalized curve.

[0030] A computer system is characterized by comprising: one or more processors, and a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method described above.

[0031] A computer-readable storage medium is characterized by storing computer-executable instructions, which, when executed, are used to implement the above-described method.

[0032] Beneficial effects

[0033] This invention provides a wide-swath SAR image intensity equalization method based on adaptive parameter adjustment. By analyzing the radar imaging power equation, a target scattering intensity variation curve corresponding to the range dimension is established. Based on this, and combined with the azimuth pattern characteristics, an azimuth intensity variation curve is fitted. This curve is then combined with the range scattering intensity variation curve, and the resulting curve is used to perform two-dimensional adjustment of the image intensity, thereby obtaining a wide-swath SAR image with balanced brightness. Compared with existing technologies, the method of this invention has the following beneficial effects in equalizing and adjusting the uneven scattering intensity variation and distortion phenomena in synthetic aperture radar images:

[0034] Synthetic Aperture Radar (SAR) utilizes the electromagnetic field scattering characteristics to acquire images of the scattering intensity of an observed scene. When the radar observation area is large, the brightness of the image decreases significantly towards the far end due to the attenuation characteristics of electromagnetic waves, leading to a decline in the reliability of subsequent SAR image recognition and detection. This invention starts with the radar imaging principle and, combined with the characteristics of the radar system, analyzes the reasons for the attenuation of scattering intensity in both range and azimuth dimensions. In the range dimension, by deriving and rewriting the traditional radar search equation, a relationship between image signal-to-noise ratio and range variation is established. In the azimuth dimension, the azimuth scattering intensity variation curve is fitted and established using the system's transmitting antenna pattern. Subsequently, a two-dimensional image equalization curve is used to compensate for the original image, achieving image intensity equalization in wide-swath SAR images, thereby effectively improving image quality and laying a good foundation for subsequent SAR image interpretation and detection. Attached Figure Description

[0035] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0036] Figure 1 The airborne SAR imaging geometry to which this invention applies;

[0037] Figure 2(a) shows the antenna azimuth pattern after Hamming window processing;

[0038] Figure 2(b) shows the magnified antenna azimuth pattern curve;

[0039] Figure 3 This is a block diagram illustrating the engineering implementation strategy of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0041] (1) Range scattering intensity equalization curve based on radar equation

[0042] Synthetic Aperture Radar Imaging Illumination Geometry Figure 1 As shown in the figure, the straight line represents the flight path of the carrier platform, v represents the flight speed of the carrier platform, and the radar is installed on the carrier platform to perform beam scanning observation of the ground scene on one side in a frontal side-view mode. The platform's working altitude is H, and R c Let θ be the slant distance from the phase center of the radar antenna to the center of the observation scene. a With θ e These represent the radar azimuth beamwidth and elevation beamwidth, respectively.

[0043] according to Figure 1 The imaging geometry configuration in the image can be used to calculate the viewing angle α under the radar imaging beam using geometric relationships.

[0044]

[0045] Given that the radar elevation beamwidth is θ e The minimum slant range R of the radar can be calculated by combining the radar's downward angle of view. min With the farthest slant distance R max as follows:

[0046]

[0047] Using the result of the above formula, combined with the law of cosines, the corresponding range coverage width L can be calculated. r for

[0048]

[0049] As can be seen from equations (2) and (3), the swath width of SAR imaging in the range direction is mainly related to the imaging range, platform height, and radar elevation beamwidth. With technological advancements and increasingly complex battlefield environments, SAR imaging is gradually developing towards longer ranges and wider swath widths. Since the essence of SAR images is to utilize the scattering characteristics of electromagnetic waves, in order to increase the image swath width, the radar beam needs to illuminate and cover a larger area. Due to the attenuation characteristics of electromagnetic waves, the energy intensity at the far end of the image will decrease significantly, resulting in a darker image brightness and making subsequent detection and identification difficult.

[0050] To address the phenomenon of image brightness attenuation, it is necessary to adjust and equalize the image intensity. In the range direction, considering that the radar elevation beamwidth is usually wide and the beam coverage area is usually within the main lobe of the beam, the range scattering intensity is mainly affected by the operating range. Therefore, a compensation curve can be established based on the relationship between electromagnetic wave attenuation and range to achieve image range brightness equalization.

[0051] Since radar mainly relies on the echo energy scattered by the target to achieve imaging and detection, the radar equation can quantitatively describe the relationship between the operating range, radar parameters and target scattering intensity. Therefore, we consider starting with the radar equation, combining it with the system imaging parameters, rewriting and deriving the radar equation, obtaining the corresponding SAR image intensity variation curve, and achieving image range-direction intensity equalization.

[0052] The equation for a traditional search radar is given below:

[0053]

[0054] Where R represents the radar's effective range, P t Let A be the peak transmitted power of the radar, A be the antenna gain area, and η be the antenna gain area. A σ represents antenna efficiency. rcs λ is the target scattering cross-section, τ and I represent the transmitted pulse width and the number of processing accumulation points, respectively, λ is the radar transmission wavelength, K and T0 represent the Boltzmann constant and standard room temperature, respectively, and F n Where D is the system noise figure, D0 is the detection signal-to-noise ratio, and L is the signal-to-noise ratio. s With C B These represent system loss and bandwidth correction factors, respectively.

[0055] As can be seen from equation (4), this equation establishes an analytical relationship between the signal-to-noise ratio of the scattering point and the effective distance. For the brightness of the SAR image in the range direction, its brightness change corresponds exactly to the relationship in equation (4). Therefore, by combining the SAR system parameters, equation (4) is rewritten and transformed to obtain the final SAR image brightness change adjustment curve.

[0056] Equation (4) is rewritten below, assuming that the range resolution of SAR imaging is ρ. r The azimuth resolution is ρ a The scattering coefficient of a certain observed target is σ ρ Assuming the target's scattering area is exactly one pixel, its scattering cross-section σ can be obtained. rcs The relationship between the two-dimensional imaging resolution and the resolution is as follows:

[0057] σ rcs =(ρ a ·ρ r / cosα)σ ρ (5)

[0058] In SAR imaging, range resolution is determined by the transmitted signal bandwidth B, while azimuth resolution is determined by the accumulated rotation angle Δθ. int The decision was made, and there was

[0059]

[0060] When the radar range is R, the synthetic aperture accumulation time T can be obtained. a as follows

[0061]

[0062] Assuming the system pulse repetition period is PRT, equation (7) can be rewritten as follows:

[0063]

[0064] Substituting the above equation into equation (6), we get

[0065]

[0066] Substituting equations (9) and (5) into equation (4) yields the following result.

[0067]

[0068]

[0069] Known system duty cycle ratio 占 The relationship is as follows

[0070] ratio 占 =τ / PRT (11)

[0071] And the system's peak transmit power P t With the system's average transmit power P a The following relationship exists:

[0072] P a =P t ·ratio 占 (12)

[0073] Substituting equations (11) and (12) into equation (10) and rewriting it, we can obtain the following expression for the relationship between the intensity of the scattering point and the distance:

[0074]

[0075] Assume the number of pixels in the distance dimension of the obtained SAR image is N. r The pixel distance between the center point of the image is R. cThen the range of distance variation R of the image can be calculated. img for Based on the geometric relationship, project it onto the ground distance, normalize the image distance variable, and then substitute it back into equation (13) to obtain the expression for the range scattering intensity variation curve h. r (R img )for

[0076]

[0077] (2) Azimuth intensity equalization curve based on pattern gain curve

[0078] According to the principles of SAR imaging, when the number of processing points is sufficient, the azimuth swath of a SAR image is mainly determined by the azimuth beamwidth. Figure 1 It can be seen that the radar imaging azimuth swath width L a The expression is:

[0079] L a =θ a ·R c (15)

[0080] Where, θ a R is the azimuth beamwidth. c Let be the slant range of the imaging center. As can be seen from equation (15), when the effective distance is constant, a wider azimuth beamwidth is required to increase the azimuth swath width. However, a wide azimuth beamwidth often introduces severe Doppler blurring. Therefore, in actual radar system design, the azimuth beamwidth is usually not too wide. However, during imaging processing, in order to ensure sufficient imaging resolution, the number of azimuth processing points is often large, which leads to the processing imaging range being wider than the actual azimuth beam main lobe coverage range. At the edge of the main lobe of the beam, the intensity energy of the image scattering points will also decrease, resulting in a darker image brightness and making effective observation impossible. To address this problem, we consider constructing an equalization curve using the characteristics of the antenna azimuth pattern gain curve to compensate for the image intensity.

[0081] Figure 2(a) shows a typical antenna azimuth pattern after Hamming windowing. It can be seen that the conventional pattern is a sinc function. After windowing, the sidelobes are effectively suppressed, and the main lobe gain is significantly improved. In SAR imaging, the main lobe width of the illumination coverage is typically around 3 dB. Figure 2(b) shows the magnified main lobe width. It can be seen that the main lobe gain gradually decreases with increasing beamwidth. Therefore, the scene brightness on both sides of the image azimuth also becomes darker, requiring brightness equalization. Compensation adjustment based on the antenna azimuth pattern gain curve should be considered.

[0082] Considering that there will be some error between the actual trajectory of the aircraft platform and the preset ideal trajectory in actual imaging, in order to ensure the accuracy of the subsequent Doppler center estimation, it is necessary to continuously adjust the beam pointing according to the inertial navigation parameters. Therefore, there will be a certain angle of view, and it is necessary to truncate the actual azimuth curve of the antenna based on the size of the angle of view.

[0083] First, by synthesizing the aircraft's heading velocity using the northeast-sky velocity from the inertial navigation parameters, the average flight speed v of the aircraft can be obtained.

[0084]

[0085] Among them, v e Representing East Speed, v n Representing the north velocity, θ c_err This is the deviation angle between the actual heading and the ideal route.

[0086] According to the Doppler theorem, the corresponding imaging oblique angle θ can be calculated. s for:

[0087]

[0088] Where PRF is the system pulse repetition frequency, λ represents the operating wavelength, and asin(·) represents the sinusoidal angle operation. The oblique angle θ is calculated... s Then, the angle θ of the directional pattern to be captured can be determined. pattern The range is:

[0089] θ pattern ∈[-θ s ,θ s (18)

[0090] Based on the angle range given in the above formula, the gain curve of the antenna azimuth pattern is truncated. After obtaining the corresponding intensity curve, the curve needs to be interpolated according to the number of azimuth points in the image to expand the number of points of the gain curve to be consistent with the number of azimuth points.

[0091] Common interpolation methods include sinc interpolation and linear interpolation. Considering that SAR images usually have a large number of points, in order to reduce the amount of computation and facilitate engineering implementation, we use frequency domain zero-padding for processing.

[0092] First, analyze the gain curve A. dB (θ pattern After performing amplitude conversion, we have:

[0093]

[0094] Perform a Fourier transform on the result of the above transformation, transform it to the frequency domain, pad the sequence with zeros on both sides until it matches the number of azimuth points in the image, and then perform an inverse Fourier transform to the time domain to obtain the azimuth scattering intensity variation curve h. a (N a ), where N a This represents the total number of points in the directional direction.

[0095] (3) Image two-dimensional brightness equalization compensation curve

[0096] The range scattering intensity variation curve h was obtained separately. r (R img ) and the curve of azimuth scattering intensity variation h a (N a After that, it needs to be adjusted to obtain a two-dimensional curve that can directly compensate for the image.

[0097] First, analyze the range scattering intensity variation curve h. r (R img Adjustments are made to the curve, which reflects the relationship between intensity and distance. Therefore, it needs to be inverted to achieve brightness balance. After inversion, considering that mean quantization is used for image quantization, the curve factor adjustment is normalized to effectively avoid excessively large image amplitudes after curve compensation. Simultaneously, to prevent rapid curve changes from causing large contrasting brightness differences in the compensated image, the normalized distance-to-brightness balance curve is cubed to reduce its gradient.

[0098] Curve of variation of azimuth scattering intensity h a (N a The same approach is used to process it. First, the curve is inverted, then normalized, and finally windowed or squared to obtain the azimuth brightness equalization curve.

[0099] After extending the azimuth and range brightness equalization curves to two dimensions, matrix multiplication yields the final image intensity equalization matrix S. eq (R img N a ), that is,

[0100] S eq (R img N a )=(h r (R img )·one(1,N a ))·(one(R img ,1)·h a (N a(20)

[0101] Where one(·) represents a completely one vector or matrix;

[0102] Take the absolute value of the original image that needs equalization compensation to obtain the corresponding grayscale image I(R). img N a The equilibrium matrix S obtained from equation (20) eq (R img N a ) and grayscale image I(R) img N a After multiplying, the overall brightness balance and compensation of the SAR image can be achieved.

[0103] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the scope of the technology disclosed in the present invention, and such modifications or substitutions should all be covered within the scope of protection of the present invention.

Claims

1. A wide-swath SAR image intensity equalization method based on adaptive parameter adjustment, characterized in that... The steps are as follows: Step 1: Establish the range-based scattering intensity variation curve ; in, This represents the peak power of the radar transmission. For antenna gain area, Indicates antenna efficiency. For radar transmission wavelength, Indicates the aircraft's flight speed. and Representing the Boltzmann constant and standard room temperature, respectively. The system noise figure is... and These represent the system loss and bandwidth correction factors, respectively. The viewpoint under the radar imaging beam. The pixel distance from the center point of the image. The number of pixels in the distance dimension of the SAR image. This represents the range resolution of SAR imaging. The scattering coefficient is... The range of distance variation in the image. Indicates the transmit pulse width; Step 2: Based on the characteristics of the azimuth pattern, fit the azimuth scattering intensity variation curve. ; 2a) Pre-acquire the azimuth pattern curve corresponding to the imaging mode of the radar system; 2b) Calculate the corresponding oblique angle using the inertial navigation velocity parameters and system parameters; in, The system pulse repetition frequency, This indicates the operation of taking the sine angle; 2c) Determine the range of angles of the direction map to be captured based on the oblique viewing angle. Extract the gain variation curve of the antenna pattern within the corresponding angular range; 2d) The gain variation curve is interpolated and fitted to the same length as the number of azimuth points in the image to obtain the azimuth scattering intensity variation curve. ,in This represents the total number of points in the direction of orientation; Step 3: Establish a two-dimensional compensation curve to balance the image brightness; 3a) Take the absolute value of the original image data to obtain the corresponding grayscale image. , 3b) Multiply the range scattering intensity variation curve with the azimuth scattering intensity variation curve to obtain the two-dimensional image scattering intensity equalization matrix. ; 3c) Equalize the scattering intensity matrix of the two-dimensional image compensate to the original grayscale image In the middle, that is, the two are multiplied together; to obtain the final intensity-equalized grayscale image.

2. The wide-swath SAR image intensity equalization method based on adaptive parameter adjustment according to claim 1, characterized in that... Step 3 b) also includes the range scattering intensity variation curve. Preprocessing: Invert the curve, then normalize it, and finally take the cube root of the normalized curve.

3. The wide-swath SAR image intensity equalization method based on adaptive parameter adjustment according to claim 1, characterized in that... Step 3 b) also includes the curve of the change in azimuth scattering intensity. Preprocessing: Invert the curve, then normalize it, and finally window or square root the normalized curve.

4. A computer system, characterized in that... include: One or more processors, a computer-readable storage medium for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method of claim 1.

5. A computer-readable storage medium, characterized in that... The device stores computer-executable instructions, which, when executed, are used to implement the method of claim 1.

Images