A method for filtering underwater laser backscattering

Abstract

This invention discloses an underwater laser backscattering filtering method. It utilizes a cascaded, synchronously scanning underwater laser detector with a small aperture to achieve 360° omnidirectional scanning and echo signal reception. A spline least mean square nonlinear adaptive filter is used to filter out nonlinear noise in the backscattered portion of both present and absent target signals. An adaptive filter pausing module adaptively adjusts the filter pausing threshold and determines whether a target peak has arrived. When a target peak arrives, the filter pauses to preserve the original trend of the target peak. This invention enables real-time acquisition of echo signals from both present and absent targets in an underwater laser detection system and provides a certain degree of anti-interference capability. While improving the accuracy of time-based identification, it also offers better backscattering filtering, making it suitable for long-range underwater laser detection.

Description

Technical Field

[0001] The invention belongs to the field of underwater laser detection, and in particular relates to an underwater laser backscatter filtering method. Background Art

[0002] Since the discovery of the underwater blue-green light-transmitting window, the blue-green laser has opened up a new way for underwater laser detection because of its advantages such as high transmittance, high resolution and identifiable target position underwater. Although the absorption of blue-green band laser by seawater has reached the lowest level, suspended particles and phytoplankton in turbid seawater will still produce strong backscattering of laser. When detecting medium and long-distance targets, the backscattered wave peak will reach the receiver before the target and superimpose on the target wave peak that arrives later. This phenomenon seriously reduces the signal-to-noise ratio of the laser detection system and makes the traditional time identification method invalid.

[0003] Under ideal conditions, the backscatter signal can be directly filtered out and the target signal can be extracted by directly "subtracting" the non-target echo from the target echo. However, the premise for the direct subtraction method to achieve good results is that the backscatter peaks in the target and non-target signals have high consistency. But in actual situations, the time jitter of the laser receiving system, slight changes in the water environment, and slight changes in the laser transmission power will cause the collected backscatter signal to have broadening, drift, fluctuation and other noise. If these noises cannot be offset, when the target's real echo signal is weak, the peak of the backscatter signal noise part will even exceed the target peak, seriously reducing the system signal-to-noise ratio.

[0004] In order to suppress the changes in backscatter signals when there is a target or not, scholars currently usually use linear adaptive filtering algorithms including the least mean square (LMS) and recursive least squares (RLS). Cheng Zao et al. proposed a backscatter filtering algorithm based on a variable forgetting factor RLS and verified its effectiveness (Cheng Z, Yang K, Han J, et al. Improved time-of-flight range acquisition technique in underwater lidar experiments [J]. Applied Optics, 2015, 54 (18): 5715-5725). Tan Yayun et al. proposed a water backscatter filtering method based on a range-gated controllable and variable-step-size minimum mean square adaptive filter, which improved the signal-to-noise ratio after filtering by more than 20 dB (Tan Y, Zhang H, Zha B. Underwater single beam circumferentially scanning detection system using range-gated receiver and adaptive filter [J]. Journal of Modern Optics, 2017, 64 (16): 1648-1656.). At present, some achievements have been made in the research of linear adaptive filtering algorithms for backscattering, but there are more nonlinear noises for laser receiving circuits, and the backscatter filtering effect of the linear filtering algorithm cannot meet the increasing demand for underwater long-distance laser detection. Summary of the invention

[0005] The purpose of the present invention is to provide an underwater laser backscatter filtering method, which can filter out the water body backscattering peak without changing the original trend of the target peak and improve the underwater laser detection distance.

[0006] The technical solution to achieve the purpose of the present invention is: an underwater laser backscatter filtering method, the steps are as follows:

[0007] Step 1: Use a small-opening cascade synchronous scanning underwater laser detector to perform circumferential scanning detection underwater:

[0008] The small-opening cascade synchronous scanning underwater laser detector includes four quadrant detection windows, which are numbered clockwise as the first quadrant, the second quadrant, the third quadrant, and the fourth quadrant, wherein the first quadrant and the third quadrant are collectively referred to as the first detection pair, and the second quadrant and the fourth quadrant are collectively referred to as the second detection pair; in order to maintain sufficient laser emission power, only one detection pair is used for each detection operation, and the two detection pairs work alternately to achieve 360° full-circle scanning of the laser; when one quadrant of the working detection pair scans the target, the target reflects the laser to obtain an echo beam, and the small-opening cascade synchronous scanning underwater laser detector receives the echo beam to generate a target echo signal d(n); at the same time, the other quadrant of the working detection pair receives a target-free echo signal x(n).

[0009] Step 2, designing a spline minimum mean square nonlinear adaptive filter;

[0010] Step 3: Directly subtract d(n) from x(n) to obtain the difference signal between the target echo and the target echo. d(n) is used as the expected signal of the spline minimum mean square nonlinear adaptive filter, x(n) is used as the input signal of the spline minimum mean square nonlinear adaptive filter, d(n) is subtracted from the output signal y(n) of the spline minimum mean square nonlinear adaptive filter to obtain the estimated signal e(n) of the spline minimum mean square nonlinear adaptive filter, and d(n), x(n), e(n), Input adaptive filter pause block.

[0011] Step 4: The adaptive filter pause module determines whether the target peak has arrived by using the threshold detection method:

[0012] when Less than the judgment threshold e th When the target peak is considered not to have arrived, the spline minimum mean square nonlinear adaptive filter continues to work to filter out the error between x(n) and d(n) in the backscattered peak part. The final output value of the spline minimum mean square nonlinear adaptive filter is Take the estimated signal e(n) of the filter.

[0013] when Not less than the judgment threshold e th When , it is considered that the target peak has arrived. At this time, the spline minimum mean square nonlinear adaptive filter is paused, that is, its learning step is set to 0 to retain the original trend of the target peak. The final output value of the spline minimum mean square nonlinear adaptive filter is Pick

[0014] Compared with the prior art, the present invention has the following significant advantages:

[0015] (1) The present invention adopts a small-opening cascade synchronous scanning underwater laser detector and proposes a corresponding scanning method, which realizes the real-time acquisition of target echo signals with and without targets, thereby improving the algorithm-carrying implementation capability of the underwater laser detection system and improving the anti-interference capability.

[0016] (2) The present invention adopts a spline minimum mean square nonlinear adaptive filter to filter out the nonlinear noise of the backscattered part in the target and non-target signals. It has a better filtering effect on the underwater laser backscattered peak, and the signal-to-noise ratio before and after filtering can be improved by more than 31dB.

[0017] (3) The present invention adopts an adaptive filter pause module, which is used to adaptively adjust the filter pause threshold according to the characteristics of the underwater laser echo signal, so as to better retain the original trend of the target peak, thereby improving the moment identification accuracy of the underwater laser detection system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 The figure is a principle block diagram of the underwater laser backscatter filtering method of the present invention.

[0019] Figure 2 It is a schematic diagram of the laser scanning method of the small-opening cascade synchronous scanning underwater laser detector of the present invention.

[0020] Figure 3 It is a principle block diagram of the spline minimum mean square nonlinear adaptive filter of the present invention.

[0021] Figure 4 This is a principle block diagram of implementing filter pause in the present invention.

[0022] Figure 5 The present invention is a block diagram of the principle of implementing adaptive adjustment of the filter pause threshold.

[0023] Figure 6 This is a principle block diagram of the present invention for realizing underwater laser detection target recognition and anti-interference.

[0024] Figure 7 This is a diagram showing the filtering effect of an original echo signal with a signal-to-noise ratio of 6.63 dB in an embodiment of the present invention.

[0025] Figure 8 This is a diagram showing the filtering effect of an original echo signal with a signal-to-noise ratio of 0.91 dB in an embodiment of the present invention.

[0026] Fig. 9 This is a diagram showing the filtering effect of an original echo signal with a signal-to-noise ratio of -5.97 dB in an embodiment of the present invention.

[0027] Fig.10 This is a diagram showing the filtering effect of an original echo signal with a signal-to-noise ratio of -8.93dB in an embodiment of the present invention.

[0028] Fig.11 This is a diagram showing the improvement of the signal-to-noise ratio of various original echo signals before and after filtering in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] The following will further introduce the specific implementation method, as well as the technical difficulties and inventive points of this invention in combination with this design example.

[0031] Combination Figure 1 to Figure 6 , an underwater laser backscatter filtering method, the specific implementation steps are as follows:

[0032] Step 1: Use a small-opening cascade synchronous scanning underwater laser detector to perform circumferential scanning detection underwater:

[0033] Combination Figure 1 , Figure 2 The small-opening cascade synchronous scanning underwater laser detector contains four quadrant detection windows, which are numbered clockwise as the first quadrant, the second quadrant, the third quadrant, and the fourth quadrant, wherein the first quadrant and the third quadrant are collectively referred to as the first detection pair, and the second quadrant and the fourth quadrant are collectively referred to as the second detection pair. In order to maintain sufficient laser emission power, only one detection pair is working each time a detection operation is performed, and the two detection pairs work alternately to achieve 360° full-circle scanning of the laser. When one quadrant of the working detection pair scans the target, the target reflects the laser to obtain an echo beam, and the small-opening cascade synchronous scanning underwater laser detector receives the echo beam to generate a target echo signal d(n); at the same time, the other quadrant of the working detection pair receives a target-free echo signal x(n).

[0034] Combination Figure 2 ,The laser emission window of the small-opening cascade synchronous scanning underwater laser detector is divided into four quadrants, each of which contains a light-transmitting window and a laser emission prism.

[0035] The laser emission prism generates a fan-shaped scanning light field by rotating and scanning. In order to maintain sufficient laser emission power and real-time acquisition of target echo signals, the laser is emitted from two windows in opposite quadrants at a time, such as the first and third quadrants emitting light at the same time, followed by the second and fourth quadrants emitting light at the same time, and this reciprocating alternation realizes 360° full-circle scanning of the laser.

[0036] Taking the laser emission from the first and third quadrants as an example, if the first and third quadrants receive the target signal at the same time (or do not receive it at the same time), the target peak disappears after backscatter filtering, which means it is judged as an interference signal (or no signal). When only a single quadrant receives the target signal, the backscatter signal can be filtered out and the target peak can be extracted by filtering the backscatter noise from the target signal through real-time acquisition of the target signal and the non-target signal.

[0037] Step 2: Design a spline minimum mean square nonlinear adaptive filter, as follows:

[0038] Combination Figure 3 The spline minimum mean square nonlinear adaptive filter includes a linear adaptive filter and a memoryless nonlinear spline interpolation function based on an adaptive lookup table. The essence of nonlinear interpolation is to store the control points of the adaptive spline curve into the adaptive lookup table, and index the ideal interpolation point on the adaptive spline curve according to the output value of the linear adaptive filter as the output signal y(n) of the spline minimum mean square nonlinear adaptive filter at time n.

[0039] Step 3: Combine Figure 1 , d(n) and x(n) are directly subtracted to obtain the difference signal with and without target echo d(n) is used as the expected signal of the spline minimum mean square nonlinear adaptive filter, x(n) is used as the input signal of the spline minimum mean square nonlinear adaptive filter, d(n) is subtracted from the output signal y(n) of the spline minimum mean square nonlinear adaptive filter to obtain the estimated signal e(n) of the spline minimum mean square nonlinear adaptive filter, and d(n), x(n), e(n), Input adaptive filter pause block.

[0040] Step 4: The adaptive filter pause module determines whether the target peak has arrived by using the threshold detection method:

[0041] Combination Figure 4 ,when Less than the judgment threshold e th When the target peak is considered not to have arrived, the spline minimum mean square nonlinear adaptive filter continues to work to filter out the error between x(n) and d(n) in the backscattered peak part. The final output value of the spline minimum mean square nonlinear adaptive filter is Take the estimated signal e(n) of the filter.

[0042] when Not less than the judgment threshold e thWhen , it is considered that the target peak has arrived. At this time, the spline minimum mean square nonlinear adaptive filter is paused, that is, its learning step is set to 0 to retain the original trend of the target peak. The final output value of the spline minimum mean square nonlinear adaptive filter is Pick

[0043] The function expression is as follows:

[0044]

[0045]

[0046] Where μ w (n), μ q (n) represents the learning step of the linear adaptive filter and the adaptive lookup table at time n, which can also be called the learning step of the nth iteration.

[0047] Combination Figure 5 In order to filter out the error of the backscattered part of the target and non-target signals as much as possible and improve the signal-to-noise ratio of the underwater laser fuze. Before the target peak appears, a larger filter pause threshold can make the spline minimum mean square nonlinear adaptive filter more reliably identify the noise, quickly reduce the system error, and achieve better backscatter filtering effect. After the target signal appears, the difference between the target echo and the non-target echo becomes larger, and the difference is the target peak. Therefore, we hope to take a smaller filter pause threshold at this time to identify the target peak as much as possible to retain its original trend and improve the moment identification accuracy.

[0048] The realization idea of ​​the pause threshold of the adaptive filter of the present invention is: according to the characteristics of the backscatter signal of the water body, e th The value of should be a function that is inversely proportional to time and whose change pattern conforms to an exponential decay trend.

[0049] The expression of the adaptive filter pause threshold is as follows:

[0050] e th (n) = e min +(e max -e min ) exp[ αn] (3)

[0051] Where e th (n) represents the pause threshold of the spline least mean square nonlinear adaptive filter at time n, which can also be called the pause threshold of the nth iteration, e max With e min They represent the maximum and minimum values ​​of the pause threshold of the adaptive filter, respectively. The speed control factor α is used to control e th (n) The rate of change.

[0052] Combination Figure 6 , taking the first group of detection pairs emitting laser as an example, if the first and third quadrants of the small-opening cascade synchronous scanning underwater laser detector receive the target signal at the same time, the target peak disappears after filtering by the spline minimum mean square nonlinear adaptive filter, and it is judged as an interference signal. If the first and third quadrants of the small-opening cascade synchronous scanning underwater laser detector do not receive the target signal, there is no target peak after filtering by the spline minimum mean square nonlinear adaptive filter, and it is judged as no target. When only a single quadrant receives the target signal, the backscattered signal can be filtered out and the target peak can be extracted by real-time acquisition of target signals and non-target signals, and the backscattered noise can be filtered out from the target signal by the spline minimum mean square nonlinear adaptive filter, and it is judged as a target.

[0053] Example

[0054] The backscatter filtering method of the present invention uses original echo signals with signal-to-noise ratios of 6.63dB, 0.91dB, -5.97dB, and -8.93dB as input data. The filtering effect of the present invention is as follows: Figure 7 , 8 , 9, and 10. Figure 7 In the case shown in Figure 1, the target peak amplitude is significantly greater than the backscattered peak, and the target peak can be identified using only the peak detection method. Figure 8 , Fig. 9 , Fig.10 The target peak has no amplitude advantage over the backscatter peak. After being processed by the underwater laser backscatter filtering method of the present invention, as shown by the solid line in the figure, the backscatter peak has been filtered out and can be identified at the moment.

[0055] The signal-to-noise ratio changes before and after filtering are as follows: Fig.11 As shown, the signal-to-noise ratio is improved by more than 31dB.

Claims

1. An underwater laser backscatter filtering method, characterized in that: Here are the steps: Step 1: Use a small-opening cascade synchronous scanning underwater laser detector to perform circumferential scanning detection underwater: The small-opening cascade synchronous scanning underwater laser detector includes four quadrant detection windows, which are numbered clockwise as the first quadrant, the second quadrant, the third quadrant, and the fourth quadrant, wherein the first quadrant and the third quadrant are collectively referred to as the first detection pair, and the second quadrant and the fourth quadrant are collectively referred to as the second detection pair; in order to maintain sufficient laser emission power, only one detection pair is used for each detection operation, and the two detection pairs work alternately to achieve 360° full-circle scanning of the laser; when one quadrant of the working detection pair scans the target, the target reflects the laser to obtain an echo beam, and the small-opening cascade synchronous scanning underwater laser detector receives the echo beam to generate a target echo signal d(n); at the same time, another quadrant of the working detection pair receives a target-free echo signal x(n); Step 2, designing a spline minimum mean square nonlinear adaptive filter; Step 3: Directly subtract d(n) from x(n) to obtain the difference signal between the target echo and the target echo. d(n) is used as the expected signal of the spline minimum mean square nonlinear adaptive filter, x(n) is used as the input signal of the spline minimum mean square nonlinear adaptive filter, d(n) is subtracted from the output signal y(n) of the spline minimum mean square nonlinear adaptive filter to obtain the estimated signal e(n) of the spline minimum mean square nonlinear adaptive filter, and d(n), x(n), e(n), Input adaptive filter pause module; Step 4: The adaptive filter pause module determines whether the target peak has arrived by using the threshold detection method: when Less than the judgment threshold e th When the target peak is considered not to have arrived, the spline minimum mean square nonlinear adaptive filter continues to work to filter out the error between x(n) and d(n) in the backscattered peak part. The final output value of the spline minimum mean square nonlinear adaptive filter is Take the estimated signal e(n) of the filter; when Not less than the judgment threshold e th When , it is considered that the target peak has arrived. At this time, the spline minimum mean square nonlinear adaptive filter is paused, that is, its learning step is set to 0 to retain the original trend of the target peak. The final output value of the spline minimum mean square nonlinear adaptive filter is Pick Among them, the adaptive filter pause module uses the threshold detection method to determine whether the target peak has arrived. The function expression is as follows: Where μ w (n), μ q (n) respectively represent the learning step size of the linear adaptive filter and the adaptive lookup table at time n, which can also be called the learning step size of the nth iteration; Where e th (n) represents the pause threshold of the spline least mean square nonlinear adaptive filter at time n, which can also be called the pause threshold of the nth iteration, e max With e min They represent the maximum and minimum values ​​of the pause threshold of the adaptive filter, respectively. The speed control factor α is used to control e th (n) The rate of change.

2. The underwater laser backscatter filtering method according to claim 1, characterized in that: The laser emission window of the small-opening cascade synchronous scanning underwater laser detector is divided into four quadrants, each of which contains a light-transmitting window and a laser emission prism; The laser emitting prism generates a fan-shaped scanning light field by rotating and scanning. In order to maintain sufficient laser emission power and real-time acquisition of target echo signals, two windows in opposite quadrants are kept emitting lasers each time, such as the first and third quadrants emitting light at the same time, followed by the second and fourth quadrants emitting light at the same time, and this reciprocating alternation realizes 360° full-circle scanning of the laser.

3. The underwater laser backscatter filtering method according to claim 1, characterized in that: In step 2, a spline minimum mean square nonlinear adaptive filter is designed as follows: The spline minimum mean square nonlinear adaptive filter includes a linear adaptive filter and a memoryless nonlinear spline interpolation function based on an adaptive lookup table. The essence of nonlinear interpolation is to store the control points of the adaptive spline curve in the adaptive lookup table, and index the ideal interpolation point on the adaptive spline curve according to the output value of the linear adaptive filter as the output signal y(n) of the spline minimum mean square nonlinear adaptive filter at time n.

Images