DE102014107979A1 - A method of classifying a water object, device, sonar and watercraft or stationary platform - Google Patents

Abstract

The invention relates to a method for classifying a water object, in particular a ship or a submarine, the method comprising the following steps: determining an underwater sound signal by means of a hydrophone, the underwater sound signal comprising a sound signal of the water object, determining a frequency spectrum of the detected underwater sound signal Determining a cepstrum of the frequency spectrum, determining a parameter set of the cepstrum, the parameter set having, in particular, direct parameters, and determining a water object class based on the parameter set.

Description

 The invention relates to a method for classifying a water object, in particular a ship or a submarine, on the basis of determined subsonic sound signals, as well as a device for carrying out the method and a sonar and a watercraft or stationary platform.

 Watercraft such as submarines and ships are classified by frequency lines of emitted subsonic sound signals. Some of the features used for classification are automatically extracted, and in addition, an operator of the sonar gives specific characteristics recognized by him to a continuously evolving model and accordingly classifies. Thus, the classification is especially on a manually created rules.

 In addition, underwater signals emitted by water objects are characterized by their spectral lines, for example in a DEMON (DEMON = Detection of Envelope Modulation on Noise) or LOFAR (LOFAR = Low Frequency Analysis and Recording) method. For example, propeller speeds and the number of blades can be analyzed on the basis of the patterns determined by these methods. An unambiguous classification, in which the underwater sound signals emitting object can be reliably categorized, is automated or not possible according to the prior art, or only to a limited extent.

 The object of the invention is to improve the prior art.

• – Ermitteln eines Unterwasserschallsignals mittels eines Hydrophons, wobei das Unterwasserschallsignal ein Schallsignal des Wasserobjekts aufweist,
• – Bestimmen eines Frequenzspektrums des ermittelten Unterwasserschallsignals,
• – Bestimmen eines Cepstrums aus dem Frequenzspektrum,
• – Bestimmen eines Parametersatzes des Cepstrums, wobei der Parametersatz insbesondere direkte Parameter aufweist, und
• – Bestimmen einer Wasserobjektklasse anhand des Parametersatzes.

The object is achieved by a method for classifying a water object, in particular a ship or a submarine, the method comprising the following steps:

• Detecting an underwater sound signal by means of a hydrophone, wherein the underwater sound signal comprises a sound signal of the water object,
• Determining a frequency spectrum of the detected underwater sound signal,
• Determining a cepstrum from the frequency spectrum,
• Determining a parameter set of the cepstrum, the parameter set having, in particular, direct parameters, and
• - Determine a water object class based on the parameter set.

 Thus, a water object can be classified automatically based on the determined parameter set. Additionally, additional or supplemental information from a sonar operator is not needed, so the present classification is not based on a manual set of rules.

The following concept is explained:

A "classifying" means in particular the classification into a specific (vehicle) category. Since the "water objects" are in particular ships, submarines, manned or unmanned vessels or sound-emitting buoys, these can be made in certain categories here. For example, fish trawlers or small motorized vessels can form a first category, submarines (submarines) and AUVs (autonomous underwater vehicles) or ROVs (remotely operated vehicles). Another class, for example, form large container ships. The classes may be arbitrarily finely divided, the summary criteria being generally speed and / or size and / or applications.

 A "hydrophone" is in particular a sound-sensitive sensor, which converts underwater sound signals into electrical signals. The hydrophone may in this case be part of a passive or active sonar.

 An "underwater sound signal" is generally a signal emitted or reflected by the water object, wherein in the present case particularly good results are achieved with original submarine sound signals emitted by the object.

 The signals recorded by the hydrophones are recorded especially in the time domain. These underwater sound signals present in the time domain can be converted, for example, by a transformation into a frequency domain, so that a "frequency spectrum" exists in this regard. In particular, the proportions of the frequencies that form the underwater sound signal are shown in this frequency spectrum.

A "cepstrum" includes in particular a renewed transformation of the frequency spectrum. In this case, the frequency spectrum may have previously been processed, for example by means of logarithming or other mathematical influences. By forming the cepstrum, a "set of parameters" can be extracted that adequately describes the cepstrum. Parameters resulting directly from the cepstrum are referred to here as direct parameters.

 On the basis of this determined parameter set, a water object can be assigned to a water object class. This can be done, for example, by determining the water object class based on the parameter set by means of a comparison, a correlation or other comparative functions.

 A "water object class" is in particular a category into which a water object is classified.

 In one embodiment, the underwater sound signal is weighted. As a result, advantageously higher-frequency components of the underwater signal can be highlighted.

 By "weighting" is meant, in particular, that gains and / or attenuations of the signal occur in a linear or non-linear manner. The weighting can be done both in the time domain and in the frequency domain.

 In order to improve the digital signal processing, a window function, in particular a Hamming window function, may be applied to the underwater sound signal and / or the weighted underwater sound signal. In particular, this window function can be used to determine the weighting with which samples obtained during the sampling of a signal are received within a section (window) in subsequent calculations. In particular, in a frequency analysis so-called leak effects can be avoided. In the present case, the window function can be applied both in the time domain and in the frequency domain.

 In a further embodiment, the determination of the frequency spectrum is effected by means of a fast Fourier transformation. Thus, an effective method for transforming into the frequency domain can be provided. In particular, such functions can be realized by means of computers or separately designed components, such as FPGAs (Field Programmable Gate Arrays).

 In order to obtain a manageable set of parameters, determining the cepstrum may comprise logarithmizing an absolute value and / or logarithmizing a plurality of absolute values and / or logarithmizing all the absolute values of the frequency spectrum.

 "Logarithm" can be done to any base such as base 2 or base 10.

 The "absolute values" are calculated in particular by amount formation from the complex frequency spectrum

 In addition, the improvement of the parameter set serves to carry out a discrete cosine transformation when determining the cepstrum.

 The modified discrete cosine transformation is also included here. In particular, the discrete cosine transform counts among the real-valued, discrete, linear, orthogonal transforms, which generally transform a discrete-time signal from the time domain or a spatial domain into a frequency domain similar to the discrete Fourier transform.

 In order to increase the information content of the parameter set, parameters derived from the direct parameters of the parameter set can be formed, so that the parameter set includes (with) derived parameters. It should be noted at this point that, of course, only derived parameters can be used.

 A "derived parameter" may be, for example, a delta value or a difference between two direct parameters or a direct parameter with a derived parameter. The direct parameters are, in particular, real-valued numerical values.

Since the meaningfulness of the individual parameters of the parameter set can be different, the direct parameters and / or the derived parameters can be weighted. Such a weighting function may be, for example, a linear or a nonlinear function, which amplifies and / or attenuates the parameters.

 In order to realize the determination of a water object class on the basis of the parameter set and the assignment of the water object to this water object class, the water object class can be determined by means of a comparison of the parameter set with a comparison parameter set, wherein in particular a correlation and / or a detection takes place by means of a Gaussian mixed distribution and / or the comparison is performed by means of a trained neural network.

 In particular, already known water objects were assigned to an object class and the known underwater sound signals and the resulting "comparison parameter sets" for the water object class were determined.

 In particular, all individual (comparison) parameter sets that apply to a water object class can be averaged so that a comparison is made with these averaged parameters. If the determined parameter set is correlated with the averaged (comparison) parameter set of the object class, a probability is determined with which the water object of this object class is associated.

 When using the Gaussian mixing distribution or the neural network, the corresponding computers or computer models are trained by means of known water object data and thus the associated parameter sets. The more test data available, the better the water objects can be reliably assigned to the individual water objects in these implementations.

wobei K die Anzahl der gemischten Komponenten, µ_k der Mittelwert einer Komponente Σ_k, die Kovarianzmatrix und c_k die Gewichtung der Mischung einer Komponente ist. A mixed distribution is understood in particular to be a composite distribution. The principles are the same as in probability theory, so that after applying the Gaussian Mixture (GMM = Gaussian Mixture Model) the basic function is

where K is the number of mixed components, μ _{k is} the average of a component Σ _k , the covariance matrix and c _{k is} the weight of the mixture of a component.

 In particular, the GMM is trained for each individual class.

 The water object class that has reached the highest probability is assigned the corresponding water object.

 The neural networks present here are, in particular, artificial neural networks, which are simulated by means of a computer, for example. These neural networks are also trained on the individual water object classes, wherein a trained neural network, for example, in a hardware structure such as an FPGA can be transferred.

 In a further embodiment, the object is achieved by a device which is set up such that a previously described method can be carried out.

 Such a device may be, for example, a computer with corresponding input and output means. An FPGA is also included in this device.

 In another embodiment, the object is achieved by a sonar, in particular a passive sonar, which has a device described above.

 Finally, the object is achieved by a watercraft or a stationary platform, in particular ship, buoy, submarine, AUV, ROV, underwater hull, hovercraft, which or which has a previously described sonar or a device described above.

 Furthermore, the invention will be explained in more detail with reference to an embodiment. It shows

1 a schematic representation of a determined for a transmitted from a water object underwater signal probability with a determined affiliation to a (water object) class.

 A container ship is powered by propellers at sea. The propeller introduces an underwater sound signal into the seawater. This underwater sound signal is detected in the present case by a passive sonar and processed accordingly electronically.

The prepared time signal is written with the function S _n = S _i - aS _i -1, where S _{n is} the weighted signal, S _{i is} the signal at time step i, a is the weighting parameter, and S _i -1 is the signal at time step i-1. Of the Value range of a is between 0 and 1, typically close to one.

Subsequently, this weighted input signal S _{n is transferred} into two overlapping segments by means of a Hamming window. Thus, a reasonable time resolution is achieved for the downstream parameters.

 For each segment, the associated frequency spectrum is determined by means of a fast Fourier transformation.

 Subsequently, the absolute values of each individual frequency spectrum and the associated logarithms to the base 10 are formed.

 Subsequently, this logarithmized spectrum is adjusted by means of bandpass filters, so that in particular a frequency axis adapted to human frequency perception is present. The advantage here is especially in a dimensional reduction.

 Subsequently, the filtered spectra are converted by means of a discrete cosine transformation, resulting in thirteen direct parameters.

 These thirteen direct parameters are amplified by means of a linear function so that higher-order parameters experience greater amplification.

 In addition, delta parameters are formed, so that there are further (derived) parameters for the thirteen direct parameters through difference formation. In this case, both adjacent parameters and the first with the third, the second with the fourth, the third with the fifth, etc. derived parameters are formed.

In the present case four water object classes were formed. The first class includes submarines 105 The second class includes trawlers and small vessels 109 The third class includes speedboats 113 and the fourth class includes container ships 117 ,

Each for ten different members of each class 105 . 109 . 113 . 117 The known underwater sound signals were used and the associated direct parameters and derived parameters were formed. The arithmetic mean was then calculated from these parameters.

 The measured underwater sound signal and associated direct and derived parameters are correlated with the arithmetic-mean parameter sets of the associated water object classes and the probability is determined in percent.

The water object classes for submarines 105 , and the fishing cutter class 109 and the speedboat class 113 result in a lower probability (<30%) of belonging to this class ( 107 . 111 . 115 ), where the probability 119 Belonging to the water object class of container ships 117 with over 90% is very high.

This is graphically in the 1 shown, where the probability axis 103 indicates a value between 0 and 100% and on the abscissa four areas are specified, which are each associated with one of the water object classes. The respective associated probability 107 for the submarine class 105 , the probability 111 for the trawlers 109 and the probability 115 for the speed boats 113 is so small that in the present case can be ruled out that the water object that has sent the underwater sound signal, this class is associated. Due to the high probability 119 is the ship mentioned above classified as a container ship.

LIST OF REFERENCE NUMBERS

101

 bar chart

103

 Correlation coordinate in%

105

 submarine

107

Correlation result to the submarine 105

109

 Trawler

111

Correlation result to the trawler 109

113

 speedboat

115

Correlation result to speedboat 113

117

 Container Ship

119

Correlation result to the container ship 117

Claims (12)

 Method for classifying a water object, in particular a ship or a submarine, the method comprising the following steps: Detecting an underwater sound signal by means of a hydrophone, the underwater sound signal having a sound signal of the water object, Determining a frequency spectrum of the detected underwater sound signal, Determining a cepstrum from the frequency spectrum, Determining a parameter set of the cepstrum, the parameter set having, in particular, direct parameters, and - Determine a water object class based on the parameter set. A method according to claim 1, characterized in that the underwater sound signal is weighted. Method according to one of the preceding claims, characterized in that a window function, in particular a Hamming window function, is applied to the underwater sound signal or the weighted underwater sound signal. Method according to one of the preceding claims, characterized in that the determination of the frequency spectrum is effected by means of a fast Fourier transformation. Method according to one of the preceding claims, characterized in that the determination of the cepstrum comprises a logarithm of an absolute value and / or a plurality of absolute values and / or all absolute values of the frequency spectrum. Method according to one of the preceding claims, characterized in that a discrete cosine transformation is performed when determining the cepstrum. Method according to one of the preceding claims, characterized in that derived from the direct parameters of the parameter set derived parameters are formed, so that the parameter set comprises derived parameters. Method according to one of the preceding claims, characterized in that the direct parameters and / or the derived parameters are weighted. Method according to one of the preceding claims, characterized in that the determination of the water object class by means of a comparison of the parameter set with a comparison parameter set, in particular a correlation and / or detection by means of a Gaussian mixing distribution and / or the comparison by means of a trained neural network he follows. Device which is arranged such that a method according to any one of the preceding claims is feasible. Sonar comprising a device according to claim 10. Vessel or stationary platform, in particular ship, submarine, buoy, which has a sonar according to claim 11 or an apparatus according to claim 10

Images