A method for extracting and identifying echo characteristics of bottom sediment detection and a system thereof

Abstract

A method, identification method, and system for extracting and identifying echo features from bottom sediment detection are described. The method first normalizes echo signals at different angles to correct and compensate for the effects of geometric variations. The normalized echo signals at each angle are then combined to obtain a normalized envelope corresponding to the bottom sediment. Based on this, an acoustic inversion model is established for the properties of bottom sediments and the characteristic parameters of the normalized envelope. A system for detecting physical properties of bottom sediments is also designed, integrating sonar excitation, reception, data acquisition, and spatial positioning modules. This system can detect and invert physical properties of bottom sediments under complex operating conditions.

Description

Technical Field

[0001] The present invention relates to the field of sonar detection, and in particular to a method for extracting and identifying echo features of bottom sediment detection, and a system thereof. Background Art

[0002] Over the years, with rapid socioeconomic development, large quantities of pollutants have been discharged into rivers and lakes, gradually forming layers of black, foul-smelling contaminated sediment at their bottoms. The long-term accumulation of contaminated sediments not only reduces the reservoir capacity of lakes and the flood-carrying capacity of rivers, but also causes significant water environmental problems and degrades aquatic ecosystems, representing a typical endogenous source of pollution. In water ecological restoration projects, understanding the physical properties of surface sediments in rivers and lakes is crucial to better guide the design and implementation of subsequent environmental governance and ecological restoration projects.

[0003] Determining the physical parameters of bottom mud sediments has long been a challenging task in underwater bottom surveying. Currently, the main methods for obtaining physical parameters of underwater mud include static penetration sounding, borehole sampling, acoustic detection, and radiometric measurement. Mechanical sampling methods are costly and inefficient. Furthermore, the original state of the bottom mud surface sediments can easily change during sampling and laboratory analysis, resulting in significant deviations in measurement results. Static penetration sounding and borehole sampling are limited to single-point measurements and cannot provide continuous measurements, resulting in significant disadvantages in efficiency and operating costs. Acoustic methods, on the other hand, primarily use acoustic transducers to transmit ultrasonic pulses to the bottom of the water. These pulses are reflected back to a receiving transducer upon encountering interfaces with sudden changes in acoustic impedance (i.e., the interface between different media layers on the bottom). The received acoustic signals can be used to analyze the structure of the bottom sediment layer, making this the most widely used detection method in underwater exploration and surveys. This method also facilitates continuous underway surveying. The detection mechanism of bottom mud sediments is the backscattering model of sound waves. The lower the frequency of the sound wave, the stronger its penetration ability, but the weaker its spatial resolution ability; the higher the frequency, the weaker its penetration ability, but the better its spatial resolution ability.

[0004] At present, acoustic detection of bottom sediments in water bodies mainly focuses on marine water bodies, and acoustic detection is still mainly based on the classification of bottom sediments. The sound wave frequency used is mainly below 500KHz. However, there is no mature method for detecting more physical property parameters of bottom sediments in shallow rivers and lakes. Summary of the Invention

[0005] In order to solve the defects of the above-mentioned existing technologies in the detection of physical property parameters of shallow river and lake bottom mud sediments, the present invention proposes a method for extracting echo features of bottom sediment detection, which is conducive to the comprehensive expression of bottom sediment characteristics, thereby improving the accuracy of subsequent bottom sediment identification.

[0006] The present invention proposes a method for extracting echo characteristics of bottom sediment detection, comprising the following steps:

[0007] S11, obtain multiple sonar echoes for the same bottom, analyze the sonar echoes, obtain the echo signal corresponding to each sonar echo, and let p m (Δt) represents the amplitude of the sampling point in the m-th echo signal with a time interval of Δt from the pulse start time of the echo signal;

[0008] S12, Indicates p m (Δt) is the normalized value, which is used to obtain the normalized envelope of multiple echo signals corresponding to the same bottom sediment. Denoted as:

[0009]

[0010] Where M represents the number of echo signals corresponding to the same substrate;

[0011] S13. Extracting normalized envelope features to construct intra-class features of the substrate.

[0012] Preferably, in S12 Calculated according to the following formula:

[0013]

[0014] min(p m (Δt)) and max(p m (Δt)) represent curves p m Minimum and maximum values ​​in (Δt);

[0015] Let S(t) denote the G-envelope The energy accumulation curve, The intra-class features constructed in S13 include: normalized envelope The starting time t1, excitation pulse duration, end time t3 and peak time t p ; The area S enclosed by the energy accumulation curve S(t) and the amplitude axis a ; The area S enclosed by the energy accumulation curve S(t) and the time axis b .

[0016] Preferably, the multiple sonar echoes obtained for the same bottom sediment belong to multiple different detection angles.

[0017] The present invention also proposes a method for detecting and identifying bottom sediments, which realizes a one-to-one correspondence between the bottom sediment identification model and the bottom sediment category based on bottom sediment classification, making the extraction of bottom sediment characteristic parameters more accurate and reliable.

[0018] The present invention also proposes a method for detecting and identifying bottom sediments, which is used to determine the normalized envelope of the bottom sediments according to the bottom sediment detection and identification model. Processing is performed to obtain the type and characteristic parameters of the substrate to be identified, wherein the characteristic parameters include one or more of bulk density, organic matter content, and water content;

[0019] The construction of the bottom sediment detection and identification model includes the following steps:

[0020] S21, setting the classification of the bottom sediment, and obtaining the normalized envelope corresponding to multiple different categories of bottom sediment according to the bottom sediment detection echo feature extraction method

[0021] S22, set category sample (x i ,y i ), where x i is the normalized envelope corresponding to the i-th sediment sample y i The category to which the ith substrate sample belongs; a substrate classification module is constructed based on multiple category samples. The substrate classification module is used to classify the substrate according to the normalized envelope of the substrate. Determine the category of the substrate;

[0022] S23. Constructing intra-class feature samples represents the intra-class feature corresponding to the j-th substrate sample of category c, Represents the characteristic parameters of the jth substrate sample of category c, based on multiple intra-class feature samples Constructing a substrate recognition module, which corresponds one-to-one to the classification of the substrate. The substrate recognition module is used to determine the characteristic parameters of the substrate based on the intra-class feature set of the substrate;

[0023] S24, combining the bottom sediment classification module and the bottom sediment identification module to build a bottom sediment detection and identification model, which is used to identify the bottom sediment according to the normalized envelope corresponding to the bottom sediment. Identify the category of the substrate and obtain characteristic parameters of the substrate.

[0024] Preferred: t1, t3 and t p Represent the normalized envelope The start time, end time and peak time of the time length t2-t1 represent the normalized envelope The duration of the excitation pulse; S a Represents the area enclosed by the energy accumulation curve S(t) and the amplitude axis, S b Represents the area enclosed by the energy accumulation curve S(t) and the time axis.

[0025] Preferably, the substrate classification module in step S22 is constructed based on an SVM classification method or a neural network autonomous learning method.

[0026] Preferably, the substrate identification module in step S23 is characterized by a multiple regression equation, which is expressed as:

[0027]

[0028] in, and ε are model parameters.

[0029] The present invention also proposes a bottom sediment detection and identification system, which includes a memory, in which a bottom sediment bottom detection and identification model and a computer program are stored. When the computer program is executed, it is used to implement the bottom sediment detection and identification method.

[0030] Preferably, the apparatus further includes a processor connected to the memory, the processor being used to obtain sonar echoes of the bottom sediment to be identified and to process the sonar echoes to obtain corresponding normalized envelope curves; the processor being used to execute a computer program stored in the memory to implement the method for detecting and identifying bottom sediments.

[0031] Preferably, it also includes a sonar excitation module, a sonar transducer module and a GPS module. The sonar excitation module receives the sound wave driving instruction issued by the processor and drives the sonar transducer module to emit sound waves to the bottom of the water. The sonar transducer module receives the sonar echo and sends it to the processor; the processor is used to execute the computer program stored in the memory to implement the bottom sediment detection and identification method; the GPS is used for positioning and timing during operation.

[0032] The advantages of the present invention are:

[0033] (1) The present invention proposes a method for extracting echo features from bottom sediment detection. First, the echo signals at different angles are normalized to correct and compensate for the effects of geometric changes. Then, the normalized echo signals at each angle are combined and normalized to obtain a normalized envelope corresponding to the bottom sediment. In this way, the resulting normalized envelope combines echo signals from all angles of the same bottom sediment, facilitating a comprehensive representation of bottom sediment characteristics, thereby improving the accuracy of subsequent bottom sediment identification.

[0034] (2) The echo features extracted in the present invention include the curve features of the normalized envelope line and the area features of the energy accumulation curve. The curve features of the normalized envelope line realize the display of the personalized features of the echo signal, and the area features of the energy accumulation curve further realize the denoising of the normalized envelope curve. The combination of the two makes the features take into account the comprehensiveness and accuracy of the echo signal features.

[0035] (3) In the present invention, the echo signal is normalized to achieve correction and enhancement of the echo signal, which is beneficial to improving the accuracy of signal processing and avoiding the adverse effects of accidental pulses on the extraction of bottom features.

[0036] (4) For the same bottom sediment, sonar echoes at multiple different detection angles are obtained. The normalized envelope curve finally obtained realizes the complementation of different echo angles, corrects and increases the data quality, and is conducive to the multi-directional reflection of bottom sediment characteristics, ensuring the diversity and accuracy of the normalized envelope line in reflecting the bottom sediment characteristics.

[0037] (5) The method for detecting and identifying underwater sediments provided by the present invention first identifies the sediment classification based on the normalized envelope, and then identifies the sediment characteristic parameters using a sediment recognition module corresponding to the sediment classification. Thus, based on the sediment classification, the present invention achieves a one-to-one correspondence between the sediment recognition model and the sediment classification, making the extraction of sediment characteristic parameters more accurate and reliable.

[0038] (6) In the present invention, the substrate classification module and the substrate identification module are nested to achieve classification and then identification of the substrate, which is highly efficient and produces accurate results.

[0039] (7) In the present invention, a multi-purpose regression equation is used to construct a bottom sediment identification module, which is simple to calculate, helps to improve the identification efficiency, and ensures the identification accuracy.

[0040] (8) The underwater sediment detection and identification system proposed in the present invention provides a carrier for the promotion of the above-mentioned underwater sediment detection and identification method. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a method for extracting echo features from bottom sediment detection proposed by the present invention;

[0042] Figure 2 This is a flow chart of the method for constructing a bottom sediment detection and identification model proposed in the present invention;

[0043] Figure 3 This is a flow chart of a method for detecting and identifying bottom sediments;

[0044] Figure 4 This is a schematic diagram of ultrasonic detection, where A\B represent two different echo angles;

[0045] Figure 5 for Figure 4 Echo signals at multiple different angles in the embodiment shown curve chart;

[0046] Figure 6 for Figure 5 Multiple The normalized envelope of Trend chart;

[0047] Figure 7 Schematic diagram of the relationship between the normalized envelope and the energy accumulation curve;

[0048] Figure 8 Schematic diagram of the energy accumulation curve and the area enclosed by the horizontal and vertical coordinates. DETAILED DESCRIPTION

[0049] A method for extracting echo features from bottom sediment detection

[0050] Reference Figure 1 This embodiment proposes a method for extracting echo features from underwater sediment detection. First, the echo signals at each angle are normalized to reduce the impact of noise. Then, the normalized echo signals at each angle are combined and normalized to obtain a normalized envelope corresponding to the sediment. This resulting normalized envelope combines echo signals from all angles of the same sediment, facilitating a comprehensive representation of sediment characteristics and improving the accuracy of subsequent sediment identification.

[0051] This embodiment provides a method for extracting echo features from underwater sediment detection, comprising the following steps:

[0052] S11. Acquire multiple sonar echoes for the same bottom sediment, analyze the sonar echoes, and obtain echo signals corresponding to each sonar echo.

[0053] Assume that the starting time of the mth echo signal is t s , the acquisition time of signal point A on the echo signal is t t , let Δt = t t -t s , then the amplitude of the signal point A on the mth echo signal is expressed as p m (Δt), p m (Δt) is a curve with time interval Δt as the time variable.

[0054] S12, order Indicates p m (Δt) is the normalized value, which is used to obtain the normalized envelope of multiple echo signals corresponding to the same bottom sediment. Denoted as:

[0055]

[0056] Here, M represents the number of echo signals corresponding to the same substrate. In practice, M can be a fixed value or a variable value. For example, if three sonar echoes are collected from a certain substrate, then M = 3; if five sonar echoes are collected from a different substrate, then M = 5.

[0057] S13. Extracting normalized envelope features to construct intra-class features of the substrate.

[0058] Specifically, in this embodiment, the intra-class features of the substrate are based on its normalized envelope The starting time t1, excitation pulse duration t2-t1, end time t3 and peak time t p And the area S enclosed by the energy accumulation curve S(t) and the amplitude axis a , the area S enclosed by the energy accumulation curve S(t) and the time axis b Construct. The energy accumulation curve S(t) is the normalized envelope The integral can be expressed as:

[0059] In specific implementation, the intra-class features can be set to include t2-t1, t p -t1, t3-t2, S a and S b Specifically, in this embodiment, the excitation pulse duration t2-t1 is a time length, which means that the interval between time t2 and time t1 is equal to the time length of the acoustic wave excitation wave.

[0060] When this embodiment is further implemented, p m (Δt) can be normalized according to the following formula:

[0061]

[0062] min(p m (Δt)) and max(p m (Δt)) represent curves p m The minimum and maximum values ​​in (Δt).

[0063] When this embodiment is further implemented, S11 can obtain sonar echoes in multiple different echo directions for the same bottom sediment. The normalized envelope curve finally obtained realizes the complementation at different echo angles, improves the denoising effect, and is conducive to the multi-directional reflection of bottom sediment characteristics, ensuring the diversity and accuracy of the normalized envelope line in reflecting the bottom sediment signs.

[0064] A model for detecting and identifying bottom sediments

[0065] The bottom sediment detection and identification model proposed in this embodiment includes a bottom sediment classification module and a bottom sediment identification module.

[0066] The bottom classification module is used to combine the normalized envelope Identify the substrate type.

[0067] Specifically, the sediment classification module can be trained based on known category samples, and the category samples are recorded as (x i ,y i ), where x i is the normalized envelope corresponding to the i-th sediment sample y i is the category to which the i-th substrate sample belongs.

[0068] The sediment classification module can be set to:

[0069] y i =f(θ1,x i ) (3)

[0070] Among them, f represents the model, θ1 represents the model parameters,

[0071] When training the sediment classification module, the basic model f is first determined. The basic model can be constructed using the SVM classification method, neural network autonomous learning, or other methods. Then, the category samples are learned through the basic model to fix the parameters θ1 of the basic model, thereby obtaining the sediment classification module.

[0072] In this embodiment, in order to ensure accurate extraction of substrate characteristic parameters, a substrate recognition module is constructed separately for each category of substrate, and the substrate recognition module obtains the characteristic parameters of the substrate according to the intra-class features of the substrate.

[0073] make represents the intra-class feature corresponding to the j-th substrate sample of category c, Represents the characteristic parameters of the j-th substrate sample of category c. In this embodiment, the substrate identification module is set to:

[0074]

[0075] Where ff represents the model and θ2 represents the model parameters.

[0076] In this embodiment, the intra-class features corresponding to the bottom sediment samples are normalized from the envelope Specifically, in this embodiment, let t1, t2, t3 and t p Represent the normalized envelope The start time, excitation pulse duration, end time and peak time of the pulse are given by S(t); let S(t) represent the normalized envelope Energy accumulation curve, S a is the area enclosed by the energy accumulation curve S(t) and the amplitude axis, S b It is the area enclosed by the energy accumulation curve S(t) and the time axis.

[0077] The calculation formula of S(t) is as follows:

[0078]

[0079] In this embodiment, according to t1, t2, t3 and t p 、S a and S b Construct the intra-class features of the bottom sample, which can be specifically recorded as:

[0080]

[0081] in,

[0082] In specific implementation, the model ff can be specifically a multiple regression equation. In this case, formula (4) can be written as:

[0083]

[0084] and ε are model parameters, that is,

[0085] It can be seen that in the bottom sediment detection and identification model of this embodiment, the input of the model is the normalized envelope of the sample The input end of the model is connected to the input end of the bottom sediment classification module and the input end of the bottom sediment recognition module respectively, the output end of the bottom sediment classification module is connected to the input end of the bottom sediment recognition module, and the output end of the bottom sediment recognition module is the output end of the model.

[0086] A training method for bottom sediment detection and recognition model

[0087] Reference Figure 2 When training the above-mentioned bottom sediment detection and identification model, the basic models of the bottom sediment classification module and the bottom sediment identification module are first determined. The basic model of the bottom sediment classification module can be constructed based on the SVM classification method or the neural network autonomous learning method, and the basic model of the bottom sediment identification module can adopt the multivariate linear regression equation.

[0088] In this embodiment, the training of the substrate recognition module is performed based on the substrate classification, that is, the model parameters of the substrate recognition module correspond one-to-one to the substrate classification. Therefore, it is necessary to obtain the substrate classification module first.

[0089] In this embodiment, the substrate classification module is obtained through data training. For example, in this embodiment, the substrate classification is set to include "gravel", "sand" and "mud", and then N substrate samples of known categories are obtained. The N substrate samples must include the three categories of "gravel", "sand" and "mud". Then the normalized envelope of the N substrate samples is obtained. Construct category samples (x i ,y i ), where x i is the normalized envelope corresponding to the i-th sediment sample y i is the category to which the i-th substrate sample belongs. Thus, combining N category samples (x i ,y i ) performs parameter fitting on the basic model formula (3) of the bottom identification module, or makes the basic model of the bottom identification module fit the category sample (x i ,y i ) for learning, a bottom sediment recognition module with fixed parameters can be obtained.

[0090] In this embodiment, the substrate recognition module corresponds to the substrate category one by one. Taking "gravel" as an example, when obtaining the substrate recognition module of the "gravel" category, K samples are first obtained. According to the classification results of the substrate classification module, the K samples all belong to the "gravel" category substrate, and the normalized envelope of the K samples is The intra-class features and characteristic parameters are all known quantities. In this embodiment, the intra-class feature samples are constructed based on the K samples. represents the intra-class feature corresponding to the j-th substrate sample of category c, Represents the characteristic parameters of the jth bottom sample of category c; then based on K intra-class feature samples Fit formula (4) or (4.1) to obtain a fixed-parameter bottom sediment recognition module.

[0091] It is worth noting that in this embodiment, the basic model of the substrate identification module corresponding to substrate samples of different categories is the same. To be precise, the model parameters of the substrate identification module correspond one-to-one to the category, that is, each substrate category has a corresponding set of parameters for the substrate identification module.

[0092] In specific implementation, N and K can be set between 20 and 100. When the bottom sediment identification module adopts a six-variable regression equation, the minimum value of K is 6.

[0093] In this embodiment, the basis for classifying the substrate is shown in Table 1 below.

[0094] Table 1

[0095]

[0096] A method for detecting and identifying bottom sediments

[0097] Reference Figure 3 When identifying a detection object, i.e., the bottom sediment to be identified, by using a method for detecting and identifying bottom sediments proposed in this embodiment, the following steps are specifically included:

[0098] S31, obtaining multi-angle sonar echoes of the bottom to be identified, and processing the sonar echoes according to the above-mentioned bottom sediment detection echo feature extraction method to obtain the normalized envelope of the bottom to be identified

[0099] S32, obtain the above-mentioned bottom sediment detection and identification model, and convert the normalized envelope of the bottom to be identified into Input the bottom sediment detection and identification model;

[0100] S33, the bottom sediment classification module in the bottom sediment detection and identification model is based on the normalized envelope Obtaining the category of the substrate to be identified, the substrate identification module in the bottom sediment detection and identification model updates the model parameters to match the category of the substrate to be identified based on the identification result of the substrate classification module;

[0101] S34, bottom identification module extracts normalized envelope The intra-class features of the substrate are obtained and the characteristic parameters of the substrate to be identified are output.

[0102] A system for detecting and identifying underwater sediments

[0103] This embodiment proposes a bottom sediment detection and identification system, including a memory and a processor. The memory stores a bottom sediment detection and identification model and a computer program. The bottom sediment detection and identification model is as described above. When the computer program is executed, it is used to implement the above-mentioned bottom sediment detection and identification method to convert the normalized envelope corresponding to the bottom to be identified into the normalized envelope. The bottom sediment detection and identification model is input, thereby obtaining characteristic parameters of the bottom sediment to be identified through the bottom sediment detection and identification model.

[0104] In this embodiment, the processor is connected to the memory, and the processor is used to obtain the sonar echo of the bottom to be identified and process the sonar echo to obtain the corresponding normalized envelope curve The processor is used to execute the computer program stored in the memory to implement the above-mentioned bottom sediment detection and identification method.

[0105] This embodiment also includes a sonar excitation module, a sonar transducer module, and a GPS module. The sonar excitation module receives sound wave drive instructions from the processor to drive the sonar transducer module to transmit sound waves toward the bottom of the water. The sonar transducer module also has the ability to receive echoes. The sonar transducer module sends the received echo signals to the processor, which calculates the corresponding normalized envelope curve. The GPS is used for positioning and timing during operation.

[0106] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for extracting echo features from bottom sediment detection, characterized in that: The following steps are involved: S11, obtain multiple sonar echoes for the same bottom, analyze the sonar echoes, obtain the echo signals corresponding to the sonar echoes, and let p m (Δt) represents the amplitude of the sampling point in the m-th echo signal with a time interval of Δt from the pulse start time of the echo signal; S12, Indicates p m (Δt) is the normalized value, which is used to obtain the normalized envelope of multiple echo signals corresponding to the same bottom sediment. Denoted as: Where M represents the number of echo signals corresponding to the same substrate; S13. Extracting normalized envelope The characteristics of the substrate are used to construct the intra-class characteristics of the substrate; make Indicates the intra-class features corresponding to the j-th substrate sample of category c; t1, t3 and t p Represent the normalized envelope The start time, end time and peak time of the time length t2-t1 represent the normalized envelope The excitation pulse duration; S a Represents the normalized envelope The area enclosed by the energy accumulation curve S(t) and the amplitude axis, S b Represents the area enclosed by the energy accumulation curve S(t) and the time axis; All are transitional items.

2. The method for extracting echo characteristics of bottom sediment detection according to claim 1, wherein: S12 Calculated according to the following formula: min(p m (Δt)) and max(p m (Δt)) represent curves p m Minimum and maximum values ​​in (Δt); Let S(t) denote the normalized envelope The energy accumulation curve, The intra-class features constructed in S13 include: normalized envelope The starting time t1, excitation pulse duration, end time t3 and peak time t p ; The area S enclosed by the energy accumulation curve S(t) and the amplitude axis a ; The area S enclosed by the energy accumulation curve S(t) and the time axis b .

3. The method for extracting echo characteristics of bottom sediment detection according to claim 1, wherein: Multiple sonar echoes obtained for the same bottom sediment belong to different detection angles.

4. A method for detecting and identifying underwater sediments, characterized in that: Used to identify the normalized envelope corresponding to the bottom sediment detection and identification model Processing is performed to obtain the type and characteristic parameters of the substrate to be identified, wherein the characteristic parameters include one or more of bulk density, organic matter content, and water content; The construction of the bottom sediment detection and identification model includes the following steps: S21, setting the classification of the bottom sediment, and obtaining the normalized envelope corresponding to multiple different types of bottom sediments according to the bottom sediment detection echo feature extraction method according to claim 1, 2 or 3 S22, set category sample (x i ,y i ), where x i is the normalized envelope corresponding to the i-th sediment sample y i The category to which the ith substrate sample belongs; a substrate classification module is constructed based on multiple category samples. The substrate classification module is used to classify the substrate according to the normalized envelope of the substrate. Determine the category of the substrate; S23. Constructing intra-class feature samples represents the intra-class feature corresponding to the j-th substrate sample of category c, Represents the characteristic parameters of the jth substrate sample of category c, based on multiple intra-class feature samples Constructing a substrate recognition module, which corresponds one-to-one to the classification of the substrate. The substrate recognition module is used to determine the characteristic parameters of the substrate based on the intra-class feature set of the substrate; S24, combining the bottom sediment classification module and the bottom sediment identification module to build a bottom sediment detection and identification model, which is used to identify the bottom sediment according to the normalized envelope corresponding to the bottom sediment. Identify the category of the substrate and obtain characteristic parameters of the substrate.

5. The method for detecting and identifying underwater sediments according to claim 4, wherein: The substrate classification module in step S22 is constructed based on the SVM classification method or the neural network autonomous learning method.

6. The method for detecting and identifying underwater sediments according to claim 4, wherein: The substrate identification module in step S23 is characterized by a multiple regression equation, which is expressed as: in, and ε are model parameters.

7. A system for detecting and identifying underwater sediments, characterized in that: The invention comprises a memory, wherein a bottom sediment detection and identification model and a computer program are stored in the memory. When the computer program is executed, it is used to implement the bottom sediment detection and identification method according to claim 4, 5 or 6.

8. The underwater sediment detection and identification system according to claim 7, characterized in that: The apparatus further includes a processor connected to the memory, the processor being configured to obtain sonar echoes of the bottom sediment to be identified and to process the sonar echoes to obtain corresponding normalized envelope curves; the processor being configured to execute a computer program stored in the memory to implement the method for detecting and identifying bottom sediments as described in claim 5.

9. The underwater sediment detection and identification system according to claim 8, characterized in that: It also includes a sonar excitation module, a sonar transducer module and a GPS module. The sonar excitation module receives the sound wave driving instruction issued by the processor and drives the sonar transducer module to emit sound waves to the bottom of the water. The sonar transducer module receives the sonar echo and sends it to the processor; the processor is used to execute the computer program stored in the memory to implement the bottom sediment detection and identification method as described in claim 4, 5 or 6; the GPS is used for positioning and timing during operation.

Images