A Method and System for Underwater Target Echo Detection Based on Amplitude Fluctuation Characteristics

Abstract

This invention provides an underwater target echo detection method and system based on amplitude fluctuation characteristics, comprising: Step S1: establishing a target detection model according to preset rules; Step S2: acquiring an echo image, performing sliding window processing on the waveform data in the echo image, and establishing a probability density function model; Step S3: calculating the likelihood ratio function of the waveform data based on the target detection model and the probability density function model; Step S4: calculating a threshold based on the probability density function model and the likelihood ratio function; Step S5: updating the echo image using the threshold as a criterion to obtain the target detection result. This invention provides a dynamic threshold detection method that adapts to target intensity fluctuations and background changes through constant false alarm rate detection and the Neyman-Pearson criterion, detecting echo images and forming new echo images.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustics, and more specifically, to a method and system for detecting underwater target echoes based on amplitude fluctuation characteristics. Background Technology

[0002] With the development of science and technology, the competition between sonar systems and underwater targets is constantly escalating. By altering the target's geometry and size, applying stealth materials, and improving maneuverability, a series of new underwater targets have been designed and applied in battlefield environments. As a key component of sonar, the Constant False Alarm Rate (CFAR) detector is frequently used in important civilian and military systems. The detection performance of CFAR has attracted widespread attention because its detection level directly affects whether sonar can accurately capture target signals. The detection performance of CFAR is related to the noise interference within its left and right reference windows. Classical CFAR detection methods mainly include cell averaging, large cell averaging, small cell averaging, and selection order statistics. However, in real-world environments, the non-uniformity of clutter or reverberation, the number of interfering targets, and the location information of reverberation edges are difficult to obtain, increasing the difficulty of target detection. Furthermore, when processing multiple frames of images, the different echo amplitudes of targets at different angles can easily lead to target loss under a fixed threshold. To address this, adaptive CFAR has been proposed. This method estimates different parameters based on the environment and the target, and then adjusts the detection threshold to ensure a constant false alarm rate.

[0003] Chinese patent document CN118191810A discloses a parameter-environment-adaptive constant false alarm rate (CFAR) detector based on a correlation vector machine (CVM). The detector includes: a matched filter for matching-filtering input sonar or radar signals; a square-law detector for square-law detection of the matched-filtered signal; an environment adaptation module for determining the environment of the detection unit based on data from a reference unit using a model order selection method and a CVM approach; and a parameter adaptation module for calculating a threshold based on the environment of the detection unit output by the environment adaptation module, thereby determining whether a target exists in the detection unit. The detector of this invention has better false alarm control capabilities, better environment classification capabilities than VI-CFAR, and can extract the position of reverberation edges in reverberation edge environments; it can also achieve better anti-interference capabilities than ACCA-CFAR. However, this invention adjusts the threshold based on complex backgrounds, but its performance is poor for targets with fluctuating characteristics.

[0004] From a statistical theory perspective, in multi-frame images, the random variable of target echo amplitude has different distribution types or the same distribution type but different distribution parameters. However, in existing technologies, fixed-threshold detection algorithms with known background distributions are generally used, and their detection performance analysis and system design are mainly aimed at traditional non-undulating target models. When the incoming target has distribution characteristics, the sonar system designed based on the traditional target model will be mismatched with the incoming target, leading to a decrease in system performance, deterioration in detection performance, and an increase in false alarm rate. In summary, for the problem of detecting targets with fluctuating echo amplitudes in multi-frame images, this invention conducts research on fluctuating target model parameter estimation, theoretical performance analysis of traditional detectors to guide system design that can match the system parameters of the incoming target, and detector design based on target knowledge. It presents an underwater target echo detection technology method based on amplitude fluctuation characteristics, which is of great significance for improving the accuracy of target models, the reliability of system design, and target detection capabilities. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method and system for underwater target echo detection based on amplitude fluctuation characteristics.

[0006] A method for underwater target echo detection based on amplitude fluctuation characteristics provided by the present invention includes:

[0007] Step S1: Establish a target detection model according to preset rules;

[0008] Step S2: Acquire the echo image, perform sliding window processing on the waveform data in the echo image, and establish a probability density function model;

[0009] Step S3: Calculate the likelihood ratio function of the waveform data based on the target detection model and the probability density function model;

[0010] Step S4: Calculate the threshold based on the probability density function model and the likelihood ratio function;

[0011] Step S5: Update the echo image using a threshold as a criterion to obtain the target detection result.

[0012] Preferably, step S1 includes:

[0013] Step S1.1: Preset the target detection rules as follows:

[0014]

[0015] Where H0 represents no target, H1 represents a target; y(t) represents the echo signal, s(t) represents the target echo signal, and c(t) represents noise reverberation;

[0016] Step S1.2: Generate target probability rules based on target detection rules, and set target probability determination rules at the same time;

[0017] In H i Under the condition that it is true, determine H j The probability that this is true is:

[0018] P(H j |H i )=∫p(y|H i )dy

[0019] Where i and j both represent 0 or 1, p(y|H i ) is H i The probability density function p(H) that y holds true under the given condition j |H i ) represents H i Assume that y satisfies H j The probability of;

[0020] Based on the target probability rule, the decision rule is generated as follows: p(H1|H1) and p(H0|H0) are the probabilities of a correct decision; p(H1|H0) is the false alarm probability. fa p(H0|H1) represents the probability of determining that a target exists when the target does not exist; p(H0|H1) is the false negative rate, which represents the probability of determining that no target exists when the target exists.

[0021] Step S1.3: Associate the target detection rules and the target probability determination rules to establish a target detection model.

[0022] Preferably, step S2 includes:

[0023] Step S2.1: Acquire an echo image at preset time intervals, and divide the echo image according to preset angles to obtain waveform data for multiple angle ranges;

[0024] Step S2.2: Select waveform data from any angle, combine the waveform data to form a detection unit, process the detection unit using a sliding window, and obtain the target detection unit for this task.

[0025] Specifically, in the selected waveform data, the preset sliding window length is M, M / 2 units before and after the detection unit are selected as reference units, and g adjacent units of the detection unit are selected as protection units. The detection unit, reference units and protection units are merged as the target detection unit for this time.

[0026] Step S2.3: Based on the target detection model, set the probability density calculation rules, and calculate the detection parameters of the target detection unit using the preset rules to establish the probability density model;

[0027] Specifically,

[0028] The probability density model for noise reverberation is as follows:

[0029]

[0030] c represents the random variable parameter, e is the preset parameter, where the noise follows a Gaussian distribution with mean U and variance σ. 2 ;

[0031] The probability density model of the target echo signal is:

[0032]

[0033] Among them, the amplitude fluctuation of the target echo s(t) follows a Ts distribution, and the parameters are... k represents the degrees of freedom. Let Γ(·) be the average target intensity, and let Γ(·) be the gamma function, which is:

[0034] Γ(x)=∫0 +∞ t x-1 e -t dt(x>0)

[0035] t is the integration variable.

[0036] Preferably, step S3 includes:

[0037] Step S3.1: Further process the echo data of the target detection unit and calculate the probability density of the echo signal y of the target detection unit under conditions H0 and H1 respectively;

[0038] Under condition H0:

[0039]

[0040] Under condition H1:

[0041]

[0042] Where f(c) is the probability density function of the noise, and g(yc) is the probability density of the target echo amplitude fluctuation. Step S3.2: Calculate the likelihood ratio function value:

[0043] λ(y)=p(y|H1) / p(y|H0).

[0044] Preferably, step S4 includes:

[0045] Step S4.1: Obtain the echo signal z of each reference element, calculate the corresponding λ(z) value, and calculate the probability density function p(λ(z)|H0) of λ(z) under the condition H0.

[0046] Step S4.2: Calculate the threshold μ based on the parameters of the calculated echo signal z, specifically determined by the following formula:

[0047]

[0048] Where α is the preset false alarm probability value, the threshold μ is the lower limit of the integral function, μ≥0 and has adaptive properties.

[0049] Preferably, step S5 includes:

[0050] Step S5.1: Determine the probability function value according to the Neyman-Pearson criterion, and classify the likelihood ratio function as follows:

[0051]

[0052] Step S5.2: Compare the likelihood ratio function value with the threshold value. If the likelihood ratio function is greater than μ, retain the echo signal corresponding to the target detection unit. If the likelihood ratio function is less than μ, assign a uniform preset value to the corresponding echo signal; thus obtaining the target detection result.

[0053] Step S5.3: Generate a detection echo image from the target detection results, change the sliding window position, and repeat steps S2-S5 until all target detection units have been detected, and output the final detection echo image.

[0054] An underwater target echo detection system based on amplitude fluctuation characteristics, provided by the present invention, includes:

[0055] Module M1: Establishes a target detection model based on preset rules;

[0056] Module M2: Acquires echo images, performs sliding window processing on the waveform data in the echo images, and establishes a probability density function model;

[0057] Module M3: Calculates the likelihood ratio function of waveform data based on the target detection model and the probability density function model;

[0058] Module M4: Calculates the threshold based on the probability density function model and the likelihood ratio function;

[0059] Module M5: Uses a threshold as a criterion to update the echo image and obtain the target detection result.

[0060] Preferably, module M1 includes:

[0061] Module M1.1: Preset target detection rules are as follows:

[0062]

[0063] Where H0 represents no target, H1 represents a target; y(t) represents the echo signal, s(t) represents the target echo signal, and c(t) represents noise reverberation;

[0064] Module M1.2: Generates target probability rules based on target detection rules, and sets target probability determination rules;

[0065] In H i Under the condition that it is true, determine H j The probability that this is true is:

[0066] P(H j |H i )=∫p(y|h i )dy

[0067] Where i and j both represent 0 or 1, p(y|H i ) is H i The probability density function p(H) that y holds true under the given condition j |H i ) represents H i Assume that y satisfies H j The probability of;

[0068] Based on the target probability rule, the decision rule is generated as follows: p(H1|H1) and p(H0|H0) are the probabilities of a correct decision; p(H1|H0) is the false alarm probability. fa p(H0|H1) represents the probability of determining that a target exists when the target does not exist; p(H0|H1) is the false negative rate, which represents the probability of determining that no target exists when the target exists.

[0069] Module M1.3: Links the target detection rules and target probability determination rules to establish a target detection model;

[0070] Module M2 includes:

[0071] Module M2.1: Acquires an echo image at preset intervals and divides the echo image into multiple waveform data ranges according to preset angles;

[0072] Module M2.2: Select waveform data from any angle, combine the waveform data to form a detection unit, process the detection unit in a sliding window manner, and obtain the target detection unit for this task;

[0073] Specifically, in the selected waveform data, the preset sliding window length is M, M / 2 units before and after the detection unit are selected as reference units, and g adjacent units of the detection unit are selected as protection units. The detection unit, reference units and protection units are merged as the target detection unit for this time.

[0074] Module M2.3: Based on the target detection model, the probability density calculation rules are set, and the parameters of this detection are calculated from the echo data of the target detection unit according to the preset rules to establish the probability density model;

[0075] Specifically,

[0076] The probability density model for noise reverberation is as follows:

[0077]

[0078] c represents the random variable parameter, e is the preset parameter, where the noise follows a Gaussian distribution with mean U and variance σ. 2 ;

[0079] The probability density model of the target echo signal is:

[0080]

[0081] The amplitude fluctuation of the target echo s(t) follows a Ts distribution, with parameters... k represents the degrees of freedom. Let Γ(·) be the average target intensity, and let Γ(·) be the gamma function, which is:

[0082] Γ(x)=∫0 +∞ t x-1 e -t dt(x>0)

[0083] t is the integration variable.

[0084] Preferably, module M3 includes:

[0085] Module M3.1: Further processes the echo data of the target detection unit, and calculates the probability density of the echo signal y of the target detection unit under conditions H0 and H1 respectively;

[0086] Under condition H0:

[0087]

[0088] Under condition H1:

[0089]

[0090] Where f(c) is the probability density function of the noise, and g(yc) is the probability density of the fluctuations in the target echo amplitude.

[0091] Module M3.2: Calculate the likelihood ratio function value:

[0092] λ(y) = p(y|H1) / p(y|H0);

[0093] Module M4 includes:

[0094] Module M4.1: Acquire the echo signal z of each reference element, calculate the corresponding λ(z) value, and calculate the probability density function p(λ(z)|H0) of λ(z) under the condition H0;

[0095] Module M4.2: Based on the calculated parameters of the echo signal z, calculate the threshold μ, specifically determined by the following formula:

[0096]

[0097] Where α is the preset false alarm probability value, the threshold μ is the lower limit of the integral function, μ≥0 and has adaptive properties.

[0098] Preferably, module M5 includes:

[0099] Module M5.1: Determines the probability function value based on the Neyman-Pearson criterion, classifying the likelihood ratio function as follows:

[0100]

[0101] Module M5.2: Compare the likelihood ratio function value with the threshold value. If the likelihood ratio function is greater than μ, retain the echo signal corresponding to the target detection unit. If the likelihood ratio function is less than μ, assign a uniform preset value to the corresponding echo signal; thus obtaining the target detection result.

[0102] Module M5.3: Forms the target detection results into a detection echo image, changes the sliding window position, and repeatedly triggers modules M2-M5 until all target detection units have been detected, and outputs the final detected echo image.

[0103] Compared with the prior art, the present invention has the following beneficial effects:

[0104] 1. This invention provides a dynamic threshold detection method that adapts to target intensity fluctuations and background changes by using constant false alarm rate detection and the Neyman-Pearson criterion. The effectiveness of the method is demonstrated by detecting echo images. In this invention, the threshold changes continuously, which can ensure a constant false alarm rate during detection.

[0105] 2. This invention, through a target detection model based on amplitude fluctuation characteristics, can more accurately identify the difference between underwater target echo signals and noise reverberation. By using the likelihood ratio function and threshold as detection criteria, it can significantly reduce the false alarm probability and false negative rate, and improve the accuracy of target detection.

[0106] 3. The threshold value set in this invention is adaptive, dynamically adjusting according to the echo signal parameters of each reference unit, thereby adapting to target detection requirements under different environments and conditions. This adaptive performance enables the method to maintain high detection performance in various underwater environments;

[0107] 4. This invention, through sliding window processing and the establishment of a probability density function model, can efficiently process large amounts of echo data and quickly identify potential target echo signals. Furthermore, this invention can update the detection results to the echo image in real time, facilitating subsequent analysis and processing. Attached Figure Description

[0108] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0109] Figure 1 This is a flowchart of the underwater target echo detection method based on amplitude fluctuation characteristics according to the present invention;

[0110] Figure 2 This is a flowchart of the method in Embodiment 1 of the present invention;

[0111] Figure 3 This is a schematic diagram of the sliding window described in Embodiment 1 of the present invention;

[0112] Figure 4 This is a comparison of the simulated detection images described in Embodiment 2 of the present invention;

[0113] Figure 5 This is a comparison of actual data detection images as described in Embodiment 2 of the present invention. Detailed Implementation

[0114] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0115] To study underwater target echo detection techniques with amplitude fluctuation characteristics as described above, this invention provides an echo detection technique where the background follows a Gaussian distribution model, and the target amplitude fluctuation follows a Ts distribution (derived from the chi-square distribution). The method divides the time-domain waveform into reference units, detection units, and protection units using a sliding window approach. Background model parameters are estimated using the reference units, and then adaptive threshold calculations are performed using the detection units. Finally, the Neyman-Pearson criterion is used for decision-making, and the final result is given. The flowchart of the method is shown below. Figure 2This method takes underwater targets with fluctuating amplitude characteristics as the research object, and adjusts the threshold by estimating parameters for different backgrounds, thereby ensuring a constant false alarm rate. It is more reliable and adaptive in target detection and can provide a reference for improving the detection capability of sonar.

[0116] A method for underwater target echo detection based on amplitude fluctuation characteristics provided by the present invention includes:

[0117] Step S1: Establish a target detection model according to preset rules; Step S2: Acquire an echo image, perform sliding window processing on the waveform data in the echo image, and establish a probability density function model; Step S3: Calculate the likelihood ratio function of the waveform data based on the target detection model and the probability density function model; Step S4: Calculate the threshold based on the probability density function model and the likelihood ratio function; Step S5: Update the echo image using the threshold as a criterion to obtain the target detection result.

[0118] Specifically, step S1 includes: Step S1.1: Preset the target detection rules, which are:

[0119]

[0120] Where H0 represents no target, H1 represents a target; y(t) represents the echo signal, s(t) represents the target echo signal, and c(t) represents noise reverberation;

[0121] Step S1.2: Generate target probability rules based on target detection rules, and set target probability determination rules at the same time;

[0122] In H i Under the condition that it is true, determine H j The probability that this is true is:

[0123] P(H j |H i )=∫p(y|H i )dy

[0124] Where i and j both represent 0 or 1, p(y|H i ) is H i The probability density function p(H) that y holds true under the given condition j |H i ) represents H i Assume that y satisfies H j The probability of;

[0125] Based on the target probability rule, the decision rule is generated as follows: p(H1|H1) and p(H0|H0) are the probabilities of a correct decision; p(H1|H0) is the false alarm probability. fap(H0|H1) represents the probability of determining that a target exists when the target does not exist; p(H0|H1) is the false negative rate, which represents the probability of determining that no target exists when the target exists.

[0126] Step S1.3: Associate the target detection rules and the target probability determination rules to establish a target detection model.

[0127] Specifically, step S2 includes: step S2.1: acquire an echo image every preset time interval, and divide the echo image according to a preset angle to obtain waveform data of multiple angle ranges; step S2.2: select waveform data of any angle, combine the waveform data to form a detection unit, process the detection unit in a sliding window manner, and obtain the target detection unit for this time.

[0128] Specifically, in the selected waveform data, the preset sliding window length is M. M / 2 units before and after the detection unit are selected as reference units, and g adjacent units of the detection unit are selected as protection units. The detection unit, reference units, and protection units are merged to form the target detection unit for this operation. Step S2.3: Based on the target detection model, probability density calculation rules are set, and the echo data of the target detection unit for this operation is used to calculate the parameters of this detection according to the preset rules, thus establishing a probability density model.

[0129] The probability density model for noise reverberation is as follows:

[0130]

[0131] c represents the random variable parameter, e is the preset parameter, where the noise follows a Gaussian distribution with mean U and variance σ. 2 ;

[0132] The probability density model of the target echo signal is:

[0133]

[0134] The amplitude fluctuation of the target echo s(t) follows a Ts distribution, with parameters... k represents the degrees of freedom. Let Γ(·) be the average target intensity, and let Γ(·) be the gamma function, which is:

[0135] Γ(x)=∫0 +∞ t x-1 e -t dt(x>0)

[0136] t is the integration variable.

[0137] Specifically, step S3 includes: step S3.1: further process the echo data of the target detection unit, and calculate the probability density of the echo signal y of the target detection unit under conditions H0 and H1 respectively;

[0138] Under condition H0:

[0139]

[0140] Under condition H1:

[0141]

[0142] Where f(c) is the probability density function of the noise, and g(yc) is the probability density of the target echo amplitude fluctuation. Step S3.2: Calculate the likelihood ratio function value:

[0143] λ(y)=p(y|H1) / p(y|H0).

[0144] Specifically, step S4 includes:

[0145] Step S4.1: Obtain the echo signal z of each reference element, calculate the corresponding λ(z) value, and calculate the probability density function p(λ(z)|H0) of λ(z) under the condition H0.

[0146] Step S4.2: Calculate the threshold μ based on the parameters of the calculated echo signal z, specifically determined by the following formula:

[0147]

[0148] Where α is the preset false alarm probability value, the threshold μ is the lower limit of the integral function, μ≥0 and has adaptive properties.

[0149] Specifically, step S5 includes:

[0150] Step S5.1: Determine the probability function value according to the Neyman-Pearson criterion, and classify the likelihood ratio function as follows:

[0151]

[0152] Step S5.2: Compare the likelihood ratio function value with the threshold value. If the likelihood ratio function is greater than μ, retain the echo signal corresponding to the target detection unit. If the likelihood ratio function is less than μ, assign a uniform preset value to the corresponding echo signal. For example, a minimum value can be selected. When generating the echo image, this preset value is used as a background parameter. Thus, the target detection result is obtained.

[0153] Step S5.3: Generate a detection echo image from the target detection results, change the sliding window position, and repeat steps S2-S5 until all target detection units have been detected, and output the final detection echo image.

[0154] Example 1

[0155] To achieve the above objectives, the present invention adopts the following technical solution:

[0156] Step 1: Establish an object detection model

[0157] Assuming the sonar system receives the echo signal, and after processing such as frequency conversion, demodulation, matched filtering, and sampling, the target detection problem can be written as the following binary hypothesis testing problem:

[0158]

[0159] In the binary hypothesis testing process, H0 represents the absence of a target, and H1 represents the presence of a target. If we denote the echo signal as y(t), the target echo signal as s(t), and the reverberation as c(t), then when there is no target in the detection unit, the echo only contains reverberation; when there is a target, the received signal consists of the target echo and the reverberation.

[0160] Because the observed samples are random, signal detection actually uses statistical methods to process random variables, and therefore uses statistical probability to describe it:

[0161] In H i Under the assumption that it is true, determine H j The probability of this being true is:

[0162] P(H j |H i )=∫p(y|H i )dy (2)

[0163] Where p(y|H i ) is H i Assume the probability density function of y is p(H) j |H i ) represents H i Assume that y satisfies H j The probability of p(H) under the binary assumption. j |H i There are 4 possible outcomes: p(H1|H1)(detection probability, usually expressed as p) d p(H0|H0) and p(H1|H0) represent the probability of a correct decision, and p(H1|H0) is called the false alarm probability. p is generally used as the denoting factor. faThe false negative rate (p(H0|H1)) represents the probability of determining the presence of a target when the target does not exist. This leads to the target detection model.

[0164] Step 2: Acquire detection image and slide window processing

[0165] The echo image of one frame is read and divided into time-domain waveforms at different angles. Processing is performed using a sliding window approach or by extracting the entire time-domain waveform. The process is as follows: In the waveform data at the selected angle, let the sliding window length be M. Select M / 2 reference units before and after the detection unit. To prevent signal leakage from the detection unit to neighboring reference units, the adjacent g units are used as guard units, thus ensuring effective estimation of background power. A schematic diagram is shown below. Figure 3 .

[0166] Based on the detection model, this invention considers that the background reverberation c(t) follows a Gaussian distribution, is independently and identically distributed, has a mean of U, and a variance of σ. 2 The corresponding probability density (pdf) is:

[0167]

[0168] Let c denote a random variable. The noise reverberation c(t) follows a Gaussian distribution with mean U and variance σ. 2 ,

[0169] The parameters are determined by fitting the data from the actual sliding window after removing the detection and protection units. The parameters are different for different sliding windows.

[0170] The amplitude fluctuation of the target echo s(t) follows a Ts distribution with parameters as follows: k represents the degrees of freedom. For the average target intensity, its probability density function is:

[0171]

[0172] Where Γ(·) is the gamma function, defined in the real number field as:

[0173]

[0174] The parameter values ​​are calculated statistically from historical multi-frame image data or given by experts.

[0175] In summary, the Gaussian background fitting parameters for the echo data within the sliding window and the parameters for the target echo fluctuation model are obtained.

[0176] Step 3: Calculate the likelihood ratio function

[0177] From step 2, under condition H0, the echo signal y satisfies:

[0178]

[0179] For p(y|H1), since the noise and target echo are independent, the probability density function of the sum of the two in the detection model can be calculated using the convolution formula:

[0180]

[0181] Where f(c) is the probability density function of the noise, obtained from formula (3) in step 2, and g(yc) is the probability density of the target echo amplitude fluctuation, obtained from formula (4) in step 2. In summary, the likelihood ratio function value can be calculated as follows:

[0182] λ(y)=p(y|H1) / p(y|H0) (8)

[0183] The main principle behind this step is:

[0184] This invention primarily employs the Neyman-Pearson criterion for judgment. The Neyman-Pearson criterion refers to ensuring a false alarm rate p in signal detection. fa Under the condition of α, to minimize p(H0|H1), correspondingly, to minimize p d =1-p(H0|H1)=β is the maximum.

[0185] Whether the criterion has an optimal solution depends on the decision rule, which involves partitioning the observed sample space into R0 and R1. In principle, there are infinitely many ways to partition the decision domain satisfying p(H0|H1)=α, and each method corresponds to a β. Since the likelihood function is bounded, β is a function of the partitioning method and is continuously bounded in the closed set, meaning there must exist an extremum. That is, there must exist a partitioning method that maximizes β. Therefore, the Neyman-Pearson criterion has a solution.

[0186] For this constrained optimization problem, the Lagrange multiplier method is used to construct the objective function J, which is expressed in integral form, i.e.:

[0187] J=p(H0|H1)+μ[p(H1|H0)-α]=∫ R0 p(y|H1)dy+μ[∫ R0 p(y|H0)dy-α] (9)

[0188] Where μ≥0, is a Lagrange multiplier. The above equation can be reduced to:

[0189] J=μ(1-α)+∫ R0 [p(y|H1)-μp(y|H0)]dy (10)

[0190] It can be seen that the first term in the above equation is positive. Therefore, we assign the part of the integral term where the integrand is negative to region R0, and assign the remaining part to region R1. This ensures that J is minimized.

[0191] H0: p(y|H1) <p(y|H0)

[0192] H1: p(y|H1)>p(y|H0) (11)

[0193] Written in the form of a likelihood ratio test:

[0194]

[0195] The above equation provides the solution to the constrained optimization problem, which involves dividing the portion of the likelihood ratio function greater than μ into region R1 and assigning the remaining portion to region R0. In the equation, μ represents the threshold value, which is determined by calculation in step 4.

[0196] Step 4: Calculate the threshold

[0197] For each reference cell value z within the sliding window, calculate the corresponding λ(z) value, and then use equation (3) to fit and determine the parameters (U1,σ1) to obtain the probability density function p(λ(z)|H0) of λ(z) under the H0 assumption. Then, the threshold μ is determined by the following equation:

[0198]

[0199] Where α is the given false alarm probability value, and μ is the lower limit of integration in the formula, which is adaptive.

[0200] For a detection environment with constant background power, p(λ(z)|H0) remains constant, and μ can be calculated using the above formula to achieve CFAR detection. If the background power is time-varying, p(λ(z)|H0) changes accordingly. In this case, a CFAR processing method is needed to dynamically estimate the background power level, making the threshold μ adaptive. That is, to provide an integral lower limit for the above formula as p(λ(z)|H0) changes, thereby ensuring that the above formula always holds true.

[0201] Step 5: Image Detection

[0202] For the value y of a detection unit, the corresponding value λ(y) can be calculated and then compared with the corresponding threshold μ. If it is greater than μ, it is determined that there is a target, and the value of the detection unit is retained in the final image. If it is less than μ, it is determined that there is no target, and the detection unit is assigned a minimum value in the final image. Then, it is determined whether all units have been detected. If there are still undetected units, the sliding window position is changed (moved one unit to the next time step) and the process continues from step 3. If there are no undetected units, the final detected image is output.

[0203] The theoretical basis and working principle of this invention are as follows:

[0204] (1) Constant False Alarm Rate (CAL) Detection: This is an adaptive threshold detection algorithm designed to obtain predictable detection and false alarm probabilities in interference environments. It is widely used in the automatic detection of radar and sonar signals. Its function is to enable automatic detection to overcome the influence of background interference power fluctuations and maintain a constant false alarm probability. Its basic principle is to use auxiliary data from the reference cell to estimate the background interference power of the detection cell, thereby obtaining a detection threshold that adapts to changes in the background environment.

[0205] (2) The Neyman-Pearson criterion is to maximize the detection probability by adjusting the detection statistic given the false alarm probability.

[0206] This invention performs target detection on single-frame images based on constant false alarm rate (CFAR) detection and the Neyman-Pearson criterion. First, the echo image is divided into time-domain segments, and a sliding window is used to extract reference, detection, and protection units on the time-domain signal. Then, statistical analysis is performed on the reference units to fit the corresponding normal distribution parameters. Next, the likelihood ratio function is calculated by combining the estimated target distribution parameters. Finally, the estimated threshold values ​​are compared sequentially to obtain the final detected image. This invention presents a detection method where the threshold value varies with the background, providing a theoretical basis for target detection.

[0207] Example 2

[0208] (1) Data reading

[0209] Simulated image: The signal-to-noise ratio is set to 10-20dB, the signal is 1, and the noise is additive white Gaussian noise satisfying N(0,1). A set of 800*20 data points is randomly generated, and its image is as follows. Figure 4 As shown in (a).

[0210] Actual data image: Data of 8138*288 was obtained based on the acquisition results, where 288 corresponds to a 360-degree direction and 8138 corresponds to a distance of 500 meters. The corresponding image is shown below. Figure 5 As shown in (a).

[0211] (2) Fitting distribution parameters using reference units

[0212] Simulation image: The window length is set to 800, and the protection unit length is 1, meaning the entire time-domain signal is processed at once. The average of the data excluding the detection and protection units is the estimated normal distribution U value, and then the standard deviation σ is calculated.

[0213] Actual data image: The window length is set to 300, and the protection unit length is 1. That is, 300 reference units before and after the detection unit are taken as background estimation according to the position of the detection unit. Then, the average of the data excluding the detection unit and the protection unit is the estimated normal distribution U value, and then the standard deviation σ is calculated.

[0214] (3) Likelihood ratio function calculation

[0215] Assuming the value of the detection unit is y, then the denominator of the likelihood ratio function is:

[0216]

[0217] The method for calculating the molecule p(y|H1) is as follows:

[0218]

[0219] Among them, the distribution parameters of the target echo Based on the existing multi-frame image data, the value is (0.2, -30). Therefore, the likelihood ratio function can be calculated.

[0220] (4) Threshold calculation and comparison

[0221] Assuming a false alarm rate α = 0.01, the threshold can be calculated using equation (13) and the fitting parameters (U1, σ1). Then, the likelihood ratio function λ(y) corresponding to the detection unit y is compared with the threshold. If it is greater than the threshold, the detection unit value is retained; otherwise, a minimum value is assigned.

[0222] Repeat the above steps until all units have been tested. Figure 4 (b) shows the results of the simulation data after being tested by this algorithm. Figure 5 (b) shows the results of the test data after detection by this algorithm. The results show that the target features were preserved, indicating that the detection method is effective.

[0223] The present invention also provides an underwater target echo detection system based on amplitude fluctuation characteristics. The underwater target echo detection system based on amplitude fluctuation characteristics can be implemented by executing the process steps of the underwater target echo detection method based on amplitude fluctuation characteristics. That is, those skilled in the art can understand the underwater target echo detection method based on amplitude fluctuation characteristics as a preferred embodiment of the underwater target echo detection system based on amplitude fluctuation characteristics.

[0224] According to the present invention, an underwater target echo detection system based on amplitude fluctuation characteristics includes: module M1: establishing a target detection model according to preset rules; module M2: acquiring echo images, performing sliding window processing on the waveform data in the echo images, and establishing a probability density function model; module M3: calculating the likelihood ratio function of the waveform data according to the target detection model and the probability density function model; module M4: calculating a threshold value according to the probability density function model and the likelihood ratio function; and module M5: updating the echo images using the threshold value as a criterion to obtain the target detection result.

[0225] Specifically, module M1 includes: Module M1.1: Preset target detection rules, which are:

[0226]

[0227] Where H0 represents no target, and H1 represents a target; y(t) represents the echo signal, s(t) represents the target echo signal, and c(t) represents noise reverberation; Module M1.2: Generates target probability rules based on target detection rules, and simultaneously sets target probability determination rules; in H i Under the condition that it is true, determine H j The probability that this is true is:

[0228] P(H j |H i )=∫p(y|H i )dy

[0229] Where i and j both represent 0 or 1, p(y|H i ) is H i The probability density function p(H) that y holds true under the given condition j |H i ) represents H i Assume that y satisfies H j The probability of a correct judgment is calculated based on the target probability rule. The judgment rule is then generated as follows: p(H1|H1) and p(H0|H0) are the probabilities of a correct judgment; p(H1|H0) is the probability of a false alarm. fa p(H0|H1) represents the probability of determining the presence of a target when the target does not exist; p(H0|H1) is the false negative rate, representing the probability of determining the absence of a target when the target exists; Module M1.3: associates the target detection rules and the target probability determination rules to establish a target detection model;

[0230] Module M2 includes: Module M2.1: Acquires an echo image at preset time intervals and divides the echo image according to preset angles to obtain waveform data for multiple angle ranges; Module M2.2: Selects waveform data from any angle and combines the waveform data to form a detection unit. The detection unit is processed using a sliding window to obtain the target detection unit for this task. Specifically, in the selected waveform data, the preset sliding window length is M. M / 2 units before and after the detection unit are selected as reference units, and g adjacent units of the detection unit are selected as protection units. The detection unit, reference units, and protection units are merged to form the target detection unit for this task; Module M2.3: Based on the target detection model, sets probability density calculation rules and calculates the detection parameters for this task using the echo data of the target detection unit according to the preset rules, establishing a probability density model; specifically...

[0231] The probability density model for noise reverberation is as follows:

[0232]

[0233] c represents the random variable parameter, e is the preset parameter, where the noise follows a Gaussian distribution with mean U and variance σ. 2 ;

[0234] The probability density model of the target echo signal is:

[0235]

[0236] Among them, the amplitude fluctuation of the target echo s(t) follows a Ts distribution, and the parameters are... k represents the degrees of freedom. Let Γ(·) be the average target intensity, and let Γ(·) be the gamma function, which is:

[0237] Γ(x)=∫0 +∞ t x-1 e -t dt(x>0)

[0238] t is the integration variable.

[0239] Specifically, module M3 includes: Module M3.1: further processing the echo data of the target detection unit, and calculating the probability density of the echo signal y of the target detection unit under conditions H0 and H1 respectively;

[0240] Under condition H0:

[0241]

[0242] Under condition H1:

[0243]

[0244] Where f(c) is the probability density function of the noise, g(yc) is the probability density of the target echo amplitude fluctuation, and module M3.2 calculates the likelihood ratio function value.

[0245] λ(y) = p(y|H1) / p(y|H0);

[0246] Module M4 includes: Module M4.1: Acquires the echo signal z of each reference element, calculates the corresponding λ(z) value, and calculates the probability density function p(λ(z)|H0) of λ(z) under condition H0; Module M4.2: Calculates the threshold μ based on the calculated parameters of the echo signal z, specifically determined by the following formula:

[0247]

[0248] Where α is the preset false alarm probability value, the threshold μ is the lower limit of the integral function, μ≥0 and has adaptive properties.

[0249] Specifically, module M5 includes: Module M5.1: Determines the probability function value according to the Neyman-Pearson criterion, classifies the likelihood ratio function, and...

[0250]

[0251] Module M5.2: Compare the likelihood ratio function value with the threshold value. When the likelihood ratio function is greater than μ, retain the echo signal corresponding to the target detection unit. When the likelihood ratio function is less than μ, assign a uniform preset value to the corresponding echo signal; thus obtaining the target detection result. Module M5.3: Form the target detection result into a detection echo image, change the sliding window position, and repeatedly trigger modules M2-M5 until all target detection units have been detected, and output the final detected echo image.

[0252] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0253] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for detecting underwater target echoes based on amplitude fluctuation characteristics, characterized in that, include: Step S1: Establish a target detection model according to preset rules; Step S2: Acquire the echo image, perform sliding window processing on the waveform data in the echo image, and establish a probability density function model; Step S3: Calculate the likelihood ratio function of the waveform data based on the target detection model and the probability density function model; Step S4: Calculate the threshold based on the probability density function model and the likelihood ratio function; Step S5: Update the echo image using a threshold as a criterion to obtain the target detection result; Step S1 includes: Step S1.1: Preset the target detection rules as follows: in, Indicates no objective. This indicates that you have a goal; Indicates the echo signal. Indicates the target echo signal. Indicates noise reverberation; Step S1.2: Generate target probability rules based on target detection rules, and set target probability determination rules at the same time; exist Under the condition that it is established, determine The probability that this is true is: in, Both represent 0 or 1. for under conditions y The probability density function that holds true. express Assuming, y satisfy The probability of; Based on the target probability rule, the decision rule is generated as follows: and The probability of making the correct judgment; For the false alarm probability, use This represents the probability of determining that a target exists when the target does not exist. The false negative rate represents the probability of determining that there is no target when the target exists. Step S1.3: Associate the target detection rules and the target probability determination rules to establish a target detection model; Step S2 includes: Step S2.1: Acquire an echo image at preset time intervals, and divide the echo image according to preset angles to obtain waveform data for multiple angle ranges; Step S2.2: Select waveform data from any angle, combine the waveform data to form a detection unit, process the detection unit using a sliding window, and obtain the target detection unit for this task. Specifically, in the selected waveform data, the preset sliding window length is M, M / 2 units before and after the detection unit are selected as reference units, and g adjacent units of the detection unit are selected as protection units. The detection unit, reference units and protection units are merged as the target detection unit for this time. Step S2.3: Based on the target detection model, set the probability density calculation rules, and calculate the detection parameters of the target detection unit using the preset rules to establish the probability density model; Specifically, The probability density model for noise reverberation is as follows: c represents the random variable parameter, e is the preset parameter, where the noise follows a Gaussian distribution with mean U and variance . ; The probability density model of the target echo signal is: Among them, target echo The amplitude fluctuations follow a Ts distribution, with parameters , k For degrees of freedom, For average target intensity, The gamma function is: t is the integration variable.

2. The underwater target echo detection method based on amplitude fluctuation characteristics according to claim 1, characterized in that, Step S3 includes: Step S3.1: Further process the echo data from the target detection unit, respectively in... , Under the given conditions, calculate the probability density of the echo signal y of the target detection unit in this case; exist Under the conditions: exist Under the conditions: in, Let be the probability density function of the noise. The probability density of the target echo amplitude fluctuation. Step S3.2: Calculate the likelihood ratio function value: 。 3. The underwater target echo detection method based on amplitude fluctuation characteristics according to claim 1, characterized in that, Step S4 includes: Step S4.1: Acquire the echo signal of each reference element. z Calculate the corresponding Value, and calculate exist probability density function under given conditions ; Step S4.2: Based on the calculated echo signal z The parameters are used to calculate the threshold value μ, which is determined by the following formula: in, The threshold value is the preset false alarm probability value, and the threshold μ is the lower limit of the integral function. It also has adaptive properties.

4. The underwater target echo detection method based on amplitude fluctuation characteristics according to claim 1, characterized in that, Step S5 includes: Step S5.1: Determine the probability function value according to the Neyman-Pearson criterion, and classify the likelihood ratio function as follows: Step S5.2: Compare the likelihood ratio function value with the threshold value. If the likelihood ratio function is greater than μ, retain the echo signal corresponding to the target detection unit. If the likelihood ratio function is less than μ, assign a uniform preset value to the corresponding echo signal; thus obtaining the target detection result. Step S5.3: Generate a detection echo image from the target detection results, change the sliding window position, and repeat steps S2-S5 until all target detection units have been detected, and output the final detection echo image.

5. An underwater target echo detection system based on amplitude fluctuation characteristics, characterized in that, include: Module M1: Establishes a target detection model based on preset rules; Module M2: Acquires echo images, performs sliding window processing on the waveform data in the echo images, and establishes a probability density function model; Module M3: Calculates the likelihood ratio function of waveform data based on the target detection model and the probability density function model; Module M4: Calculates the threshold based on the probability density function model and the likelihood ratio function; Module M5: Uses a threshold as a criterion to update the echo image and obtain the target detection result; Module M1 includes: Module M1.1: Preset target detection rules are as follows: in, Indicates no objective. This indicates that you have a goal; Indicates the echo signal. Indicates the target echo signal. Indicates noise reverberation; Module M1.2: Generates target probability rules based on target detection rules, and sets target probability determination rules; exist Under the condition that it is established, determine The probability that this is true is: in, Both represent 0 or 1. for under conditions y The probability density function that holds true. express Assuming, y satisfy The probability of; Based on the target probability rule, the decision rule is generated as follows: and The probability of making the correct judgment; For the false alarm probability, use This represents the probability of determining that a target exists when the target does not exist. The false negative rate represents the probability of determining that there is no target when the target exists. Module M1.3: Links the target detection rules and target probability determination rules to establish a target detection model; Module M2 includes: Module M2.1: Acquires an echo image at preset intervals and divides the echo image into multiple waveform data ranges according to preset angles; Module M2.2: Select waveform data from any angle, combine the waveform data to form a detection unit, process the detection unit in a sliding window manner, and obtain the target detection unit for this task; Specifically, in the selected waveform data, the preset sliding window length is M, M / 2 units before and after the detection unit are selected as reference units, and g adjacent units of the detection unit are selected as protection units. The detection unit, reference units and protection units are merged as the target detection unit for this time. Module M2.3: Based on the target detection model, the probability density calculation rules are set, and the parameters of this detection are calculated from the echo data of the target detection unit according to the preset rules to establish the probability density model; Specifically, The probability density model for noise reverberation is as follows: c represents the random variable parameter, e is the preset parameter, where the noise follows a Gaussian distribution with mean U and variance . ; The probability density model of the target echo signal is: Among them, target echo The amplitude fluctuations follow a Ts distribution, with parameters , k For degrees of freedom, For average target intensity, The gamma function is: t is the integration variable.

6. The underwater target echo detection system based on amplitude fluctuation characteristics according to claim 5, characterized in that, Module M3 includes: Module M3.1: Further processes the echo data from the target detection unit, specifically in... , Under the given conditions, calculate the probability density of the echo signal y of the target detection unit in this case; exist Under the conditions: exist Under the conditions: in, Let be the probability density function of the noise. The probability density of the target echo amplitude fluctuation. Module M3.2: Calculate the likelihood ratio function value: ； Module M4 includes: Module M4.1: Acquires the echo signal of each reference element. z Calculate the corresponding Value, and calculate exist probability density function under given conditions ; Module M4.2: Based on the calculated echo signal z The parameters are used to calculate the threshold value μ, which is determined by the following formula: in, The threshold value is the preset false alarm probability value, and the threshold μ is the lower limit of the integral function. It also has adaptive properties.

7. The underwater target echo detection system based on amplitude fluctuation characteristics according to claim 5, characterized in that, Module M5 includes: Module M5.1: Determines the probability function value based on the Neyman-Pearson criterion, classifying the likelihood ratio function as follows: Module M5.2: Compare the likelihood ratio function value with the threshold value. If the likelihood ratio function is greater than μ, retain the echo signal corresponding to the target detection unit. If the likelihood ratio function is less than μ, assign a uniform preset value to the corresponding echo signal; thus obtaining the target detection result. Module M5.3: Forms the target detection results into a detection echo image, changes the sliding window position, and repeatedly triggers modules M2-M5 until all target detection units have been detected, and outputs the final detected echo image.

Images