Wilcoxon-SO nonparametric detection method for ship targets in SAR image based on minimum selection logic

Abstract

The application discloses a SAR image ship target Wilcoxon-SO nonparametric detection method based on selected small logic, and relates to the technical field of image detection. The detection steps mainly include the following steps: firstly, a detection unit, a reference sliding window with a front edge and a rear edge sliding window, and a detection sliding window of a protection area are set; then, the SAR ship remote sensing image is traversed pixel by pixel by using the detection unit; subsequently, the mean ratio MR of the front edge and the rear edge sliding window is calculated, and the relationship between MR and is compared to determine whether the clutter is uniform or non-uniform; the entire reference sliding window is selected for the uniform clutter, and the front edge / rear edge sliding window with a smaller sample mean is selected for the non-uniform clutter and is widened by one time; then, the detection statistic R is calculated; finally, a detection threshold is set T , and whether there is a target in the detection unit is determined according to the relationship between R and T . The application can still reliably identify the target to be detected in the face of dense multi-target interference, and the anti-shielding detection capability in the multi-target non-uniform background is significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image detection, and in particular to a SAR image ship target Wilcoxon-SO non-parametric detection method based on small selection logic. BACKGROUND

[0002] In modern marine monitoring application scenarios, synthetic aperture radar (SAR) has become the core technical support for ship target identification and tracking tasks due to its unique advantages such as all-weather operation, all-weather coverage, wide-area detection, high-resolution imaging, and cloud penetration. With the increasing frequency of global marine shipping, fishing, and other activities, the practical demand for SAR image ship target "accurate identification, rapid response, and anti-interference detection" continues to increase, whether it is in the field of port operation scheduling, busy waterway traffic supervision, or other fields of dense ship reconnaissance and marine situation awareness.

[0003] Among various technical paths for SAR image ship target detection, radar target constant false alarm rate (CFAR) detection technology has become the most widely used and effective mainstream solution because it can stably control the false alarm probability in complex background environments. Since the two-parameter CFAR detection method based on Gaussian distribution was proposed, the industry has long focused on parametric CFAR detection algorithms, whose core logic is: assuming that the sea clutter background in the SAR image follows a certain statistical distribution (such as Gaussian distribution, negative exponential distribution, Rayleigh distribution, K distribution, alpha stable distribution, G0 distribution, lognormal distribution, Gamma distribution, or mixed Rayleigh distribution), the key parameters of this distribution are estimated through clutter sample screening or rejection techniques, and finally the detection threshold is set based on these parameters to identify ship targets.

[0004] However, in actual applications, the sea clutter of SAR images is influenced by ocean environmental factors and imaging conditions such as wind disturbance, ocean current movement, and port artificial facility reflection, showing high complexity and variability. The statistical characteristics of actual sea clutter often do not match the distribution model preset by parametric CFAR. Once this "distribution mismatch" occurs, the detection performance of parametric CFAR will significantly deteriorate, and even a large number of missed targets or false clutter will occur. At the same time, parametric CFAR needs to estimate distribution parameters through complex mathematical calculations, which has the inherent defects of large calculation amount and poor real-time performance, making it difficult to adapt to the SAR data processing requirements in multi-target dense scenarios such as ports and busy waterways.

[0005] In order to break through the technical limitations of the parameter CFAR, the applicant of the present application previously attempted to use the Wilcoxon non-parametric CFAR detection technology for SAR image ship target detection. However, practice shows that the Wilcoxon non-parametric CFAR still has obvious technical shortcomings: in SAR image ship target detection, dense ship targets (i.e. multi-target scene) often appear in port, busy channel and other areas, at this time, the pixels of adjacent ships will invade the detection sliding window of the Wilcoxon non-parametric detector, forming a "shielding effect" on the ship to be detected, resulting in a significant decline in the detection capability of the ship in the non-uniform background caused by multiple targets, which seriously limits the practical value of the technology in the dense ship scene.

[0006] Therefore, how to improve the anti-shielding detection capability of the Wilcoxon non-parametric CFAR detection technology in the multi-target non-uniform background has become a key technical problem to be solved in the field of SAR image ship target detection, which is also the core improvement direction of the present application.

[0007] The information disclosed in this BACKGROUND section is only intended to increase an understanding of the general context in which the present application can be practiced. It is not admitted that this information constitutes prior art that is already known in the art. SUMMARY

[0008] In view of the above technical problems, the embodiments of the present application provide a SAR image ship target Wilcoxon-SO non-parametric detection method based on a minimum selection logic to solve the problems proposed in the above background technology.

[0009] The present application provides the following technical solutions: a SAR image ship target Wilcoxon-SO non-parametric detection method based on a minimum selection logic, comprising the following steps:

[0010] A detection sliding window is set, which includes a detection unit located in the middle, a reference sliding window surrounding the detection unit, and a protection area between the detection unit and the reference sliding window; the reference sliding window includes a front edge sliding window and a rear edge sliding window;

[0011] The detection sliding window is used to traverse the SAR collected ship target remote sensing image pixel by pixel, and the distance moved by the detection sliding window each time is equal to the width of the detection unit, that is, t pixels;

[0012] The mean ratio MR of the front edge sliding window and the rear edge sliding window is obtained;

[0013] A threshold is set MR is compared with and respectively;

[0014] If If the reference samples of the front edge sliding window and the rear edge sliding window are uniform clutter, it is determined that the reference samples of the front edge sliding window and the rear edge sliding window are uniform clutter;

[0015] If , it is determined that the reference samples of the front edge sliding window and the rear edge sliding window are non-uniform clutter, that is, the multi-target condition appears in the sliding window is detected;

[0016] If it is determined that the uniform clutter, the whole reference sliding window is selected as the reference sample for threshold calculation;

[0017] If it is determined that the non-uniform clutter, the front edge sliding window or the rear edge sliding window with smaller sample mean is selected as the reference sample for threshold calculation, and the width of the front edge sliding window or the rear edge sliding window is expanded by 1 times, that is, the width is changed to 2 q ;

[0018] The width of the front edge sliding window or the rear edge sliding window is expanded by 1 times, that is, the width is changed to 2 q After that, when the Wilcoxon-SO non-parametric detector sets the detection threshold by using the whole reference sliding window, the front edge sliding window or the rear edge sliding window, the number of reference units in the whole reference sliding window, the front edge sliding window or the rear edge sliding window is the same, that is, 2 n .

[0019] According to formula (1), the statistical quantity for detecting the ship target of the SAR image is calculated R ; formula (1) is:

[0020] ;

[0021] Wherein, is the detection sample of the detection unit, is the reference sample of the reference sliding window, u is a unit step function, m is the number of detection samples, n is the number of reference samples;

[0022] The detection threshold is set as T , if R > T , it is determined that the ship target exists in the detection unit; if R ≤ T , it is determined that the ship target does not exist in the detection unit.

[0023] It is explained that the improved SAR image ship target Wilcoxon non-parametric detector is proposed based on the "Smallest Of" logic, therefore, the method of the application is simply referred to as the Wilcoxon-SO non-parametric detector.

[0024] Preferably, the width of the detection unit is 2 t pixels, the width of the front edge sliding window and the rear edge sliding window is 2 q pixels, and the width of the protection area is 2g 1 pixel.

[0025] Preferably, the mean ratio MR is calculated according to equation (2); equation (2) is:

[0026] ;

[0027] in, and These are the mean values ​​of the reference samples in the leading edge sliding window and the trailing edge sliding window, respectively.

[0028] Preferably, the detection threshold T Based on the preset false alarm probability It is obtained by calculation using equation (3); equation (3) is:

[0029] ;

[0030] π m , n ( k ) indicates from 1 to m + n Selected from m There are 10 values ​​and their sum is 1 k The number of possible choices; n This is the number of reference samples. m It is the number of samples tested. N = m + n .

[0031] Preferably, in equation (2) The calculation formula is obtained from formula (4); formula (4) is:

[0032] .

[0033] Preferably, the reference sliding window is used to set the detection threshold; the protection unit is used to prevent the pixel energy of adjacent ship targets from leaking into the detection unit.

[0034] Preferred threshold The probability of a uniform background being incorrectly identified as non-uniform is determined based on 0.1.

[0035] The Wilcoxon-SO nonparametric detection method for ship targets in SAR images based on small selection logic provided in this invention has the following beneficial effects:

[0036] (1) Under uniform background, the Wilcoxon-SO nonparametric detector can maintain target detection performance comparable to that of the traditional Wilcoxon nonparametric detector. However, in multi-target scenarios, the traditional Wilcoxon nonparametric detector will experience a significant drop in detection capability under moderate multi-target interference. The Wilcoxon-SO nonparametric detector can still maintain good detection performance. Even when faced with extremely dense multi-target interference, it can still reliably identify the target to be detected, significantly improving the anti-masking detection capability under non-uniform multi-target background.

[0037] (2) In the case of non-uniform background (multiple targets), select the leading / trailing sliding window with a smaller sample mean and widen it to 2. q In multi-target scenarios, interfering ships can increase the mean of reference samples. Selecting a smaller target can avoid the high-mean reference area that is being interfered with, retain the pure background samples that are not being interfered with, and ensure the detection capability of the occluded target. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the detection sliding window of the Wilcoxon-SO nonparametric detector for SAR images of ship targets based on the small selection logic of the present invention.

[0039] In the diagram, the yellow area on the right is the leading edge sliding window, and the green area on the left is the trailing edge sliding window.

[0040] Figure 2 Detection probability for Wilcoxon-SO nonparametric detector P d Curve showing the variation of SCR (dB) with signal-to-noise ratio;

[0041] Figure 3 This is a schematic diagram illustrating the specific implementation process of the Wilcoxon-SO nonparametric detector. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] The following is in conjunction with the appendix Figures 1-3 The present invention will be further described in detail below, along with specific embodiments.

[0044] To address the problems mentioned in the background section, this invention provides a Wilcoxon-SO nonparametric detection method for ship targets in SAR images based on small selection logic, in order to solve the aforementioned technical problems. The specific technical solution is as follows.

[0045] (a) Detection using the Wilcoxon-SO nonparametric detection method

[0046] Steps of the Wilcoxon-SO nonparametric detection method for ship targets in SAR images based on the minimum selection logic

[0047] Step 1: Initialize the detection sliding window and traversal method

[0048] For remote sensing images acquired by spaceborne synthetic aperture radar (SAR), a Wilcoxon-SO nonparametric detector is used for traversal detection:

[0049] The step size of the detection sliding window moving along the image each time is the width of the detection unit. t 100 pixels (of which) t =2, which is the recommended parameter);

[0050] The detection sliding window consists of two parts: a leading edge sliding window (right side area) and a trailing edge sliding window (left side area), both with an initial width of [missing information]. q (in q =3, which is the recommended parameter);

[0051] A protective area with a width of [missing information] is provided between the detection unit and the reference unit (front / back sliding window). g (in g =60, which needs to be greater than the maximum length of the ship in the SAR image to prevent energy leakage from adjacent targets.

[0052] Step 2: Refer to the dynamic selection and width adjustment of the sliding window.

[0053] The Wilcoxon-SO nonparametric detector selects the reference cell used to set the detection threshold through the following logic:

[0054] When "Entire Reference Sliding Window" is selected, all reference cells of the leading edge sliding window and the trailing edge sliding window are used directly;

[0055] When selecting "Front Edge Sliding Window" or "Rear Edge Sliding Window", its width needs to be increased by 1 to 2 times. q This ensures that regardless of which reference window is selected, the number of reference cells it contains is always the same. n (To ensure statistical consistency).

[0056] Step 3: Calculate the mean ratio (MR) and determine clutter uniformity.

[0057] Define the mean ratio (MR): Calculate the mean of the reference sample in the leading-edge sliding window. Mean of the sliding window reference sample with the trailing edge The ratio of the two is the mean ratio:

[0058] ;

[0059] Determine clutter uniformity: Compare MR with threshold and its reciprocal Comparison (where) =1.143, determined through simulation, satisfying the condition that "the probability of a uniform background being misclassified as non-uniform is 0.1").

[0060] like The reference samples for the leading and trailing edges of the sliding window are determined to be a uniform background.

[0061] like or It is determined to be a non-uniform background (i.e., a situation with "multiple targets").

[0062] Step 4: Select a reference sliding window based on clutter type

[0063] Uniform background: The detection threshold is set using a reference cell with a "whole reference sliding window";

[0064] Non-uniform background ("multi-target" case): Select the leading edge or trailing edge sliding window with a smaller mean (i.e., "select small (SO)" logic) to set the detection threshold.

[0065] Step 5: Calculate the detection statistic R And determine whether the target exists.

[0066] Define detection statistics R Let the sample of the detection unit be... x 1, x 2,…, The sample of the reference unit is y 1, y 2,…, Statistic R The total number of times the detected sample was greater than the reference sample:

[0067] ;

[0068] Determine the detection threshold T And determine: threshold T Based on the set false alarm probability The solution is obtained using the following formula:

[0069] ;

[0070] in, Indicates from 1 tom + n Selected from m There are 10 values ​​and their sum is 1 k The number of possible choices; n This is the number of reference samples. m It is the number of samples tested. N = m + n .

[0071] like R > T The detection unit was found to contain a ship target.

[0072] like R ≤ T The system determined that there were no ship targets in the detection unit.

[0073] Additional explanation: Modeling parameters for "multi-objective" scenarios

[0074] In this method, the SAR image clutter background and the ship target scene are modeled based on the following parameters (used to verify the method's performance):

[0075] Clutter background: conforms to shape parameters c =1.2 Weibull distribution;

[0076] Ship targets: Set as Swerling II type targets;

[0077] Radar echo: undergoes square law detection processing.

[0078] (ii) Detection using the Wilcoxon nonparametric detection method

[0079] The conventional Wilcoxon nonparametric detection method was used for detection; it shared the same experimental scenario as the Wilcoxon-SO nonparametric detection method: the clutter background was modeled as a Weibull distribution, the target model was Swerling type II, and square law detection was considered.

[0080] Furthermore, the parameters of both the Wilcoxon nonparametric detector and the Wilcoxon-SO nonparametric detector are set to... t =2, q =3 and g =60. Among them, Figure 2 The detection probability of the Wilcoxon-SO nonparametric detector is given. P d The curve showing the variation of the signal-to-noise ratio (SCR) (dB). To demonstrate the detector's ability to contain interfering targets, we consider a strong target interference scenario, i.e., an interference ratio (ICR) of 30dB, which is an extreme case of multi-target interference. Figure 2In the diagram, I1 represents the number of pixels of interfering ship targets entering the leading edge sliding window, and I2 represents the number of pixels of interfering ship targets entering the trailing edge sliding window. I1=0, I2=0 indicates a uniform background.

[0081] (III) Test Results

[0082] Figure 2 It can be seen that in a uniform background, the Wilcoxon-SO nonparametric detector maintains the same target detection capability as the Wilcoxon nonparametric detector, meaning their detection probability curves almost overlap. When I1=10 and I2=20, i.e., the total number of interfering pixels is 30, the Wilcoxon nonparametric detector still has the ability to detect interfering targets. However, when I1=20 and I2=40, i.e., the total number of interfering pixels is 60, the Wilcoxon nonparametric detector almost loses its ability to detect targets with SCR=20dB. However, the Wilcoxon-SO nonparametric detector still has good detection capability when I1=20 and I2=40 (total number of interfering pixels is 60), and even when I1=30 and I2=300 (total number of interfering pixels is 330), the Wilcoxon-SO nonparametric detector can still reliably detect the target.

[0083] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A Wilcoxon-SO nonparametric detection method for ship targets in SAR images based on selected small logic, characterized in that, The method comprises the following steps: setting a detection sliding window, which comprises a detection unit in the middle, a reference sliding window surrounding the detection unit, and a protection area between the detection unit and the reference sliding window; the reference sliding window comprises a front sliding window and a rear sliding window; using the detection sliding window to traverse the SAR collected ship target remote sensing image pixel by pixel, the distance of each movement of the detection sliding window being equal to the width of the detection unit; obtaining the mean ratio MR of the front sliding window and the rear sliding window; Setting threshold K MR MR is compared with K MR and K -1 MR respectively; If K -1 MR ≤ MR≤ K MR , then determine that the reference samples of the leading and trailing sliding windows are uniform clutter. If MR < K -1 MR or MR > K MR , then it is determined that the reference samples of the front edge sliding window and the back edge sliding window are non-uniform clutter, that is, a multi-target situation occurs in the sliding window. if it is determined that the clutter is uniform, selecting the entire reference sliding window as the reference sample for threshold calculation; if it is determined that the clutter is non-uniform, selecting the front sliding window or the rear sliding window with smaller sample mean as the reference sample for threshold calculation, and expanding the width of the selected front sliding window or rear sliding window by 1 time; calculating the statistic R for detecting the ship target in the SAR image according to formula (1); Formula (1) is: ; wherein x i is a detection sample of the detection unit, y j is a reference sample of the reference sliding window, u is a unit step function, m is the number of detection samples, and n is the number of reference samples; setting a detection threshold T, if R>T, it is determined that there is a ship target in the detection unit; if R≤T, it is determined that there is no ship target in the detection unit; the mean ratio MR is calculated according to formula (2); formula (2) is: ； wherein, and are the mean values of the reference samples in the leading and trailing sliding windows, respectively. The detection threshold T is determined according to a preset false alarm probability P fa is calculated by equation (3); equation (3) is: ； π m,n (k) denotes the number of possible choices of m values from 1 to m+n such that their sum is k; n is the number of reference samples, m is the number of test samples, N = m+n; The calculation formula of the formula (3) is calculated according to the formula (4); the formula (4) is: The calculation formula of the formula (3) is calculated according to the formula (4); the formula (4) is: ； Threshold K MR Determined in accordance with the probability of 0.1 that a uniform background is erroneously classified as non-uniform.

2. The Wilcoxon-SO non-parametric detection method for ship targets in SAR images based on selected small logic according to claim 1, characterized in that, the width of the detection unit is t pixels, the width of the front sliding window and the rear sliding window is q pixels, and the width of the protection area is g pixels.

3. The Wilcoxon-SO non-parametric detection method for ship targets in SAR images based on selected small logic according to claim 1, characterized in that, The reference sliding window is used to set the detection threshold; and the protection unit is used to prevent the pixel energy of adjacent ship targets from leaking into the detection unit.

Images