Device and Method for Generating Clutter Map using AI in Clutter Environments

Abstract

An artificial intelligence-based clutter map generation device in a clutter environment according to an embodiment of the present invention comprises, in a target detection device using radar, a radar signal input unit that receives a current stage surrounding scan signal by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing unit that provides a previous stage estimation parameter based on the surrounding scan signal of the preceding stage of the obtained surrounding scan signal; a parameter generating unit that estimates a current stage parameter using the current stage surrounding scan signal and the previous stage estimation parameter; and a clutter map generating unit that generates a probability distribution for a specific clutter by repeating the parameter estimation and thereby generates a statistics-based clutter map. By doing so, it is possible to generate a clutter map regardless of interference information caused by instantaneously occurring ECM jamming, etc., and to obtain accurate results during clustering and target tracking by tagging the clutter probability for the detected signal.

Description

Technology Field

[0001] The present invention relates to an artificial intelligence-based clutter map generation device and method in a clutter environment. Background Technology

[0002] A radar system emits radar signals through radar sensors and detects desired targets by processing the received signals reflected by targets and other elements. Most radars are operated to observe specific targets, but the received signals contain unwanted signal components depending on the operating environment, in addition to the target intended for detection. These unwanted signal components are referred to as radar clutter.

[0003] In general, the detection performance of radar sensors is significantly affected by clutter components. In cluttered environments, removing clutter components (i.e., clutter signals) is essential and is the most critical factor in the design of radar systems and the detection of moving targets.

[0004] In the design of radar systems and the detection of moving targets, the radar system uses a clutter map technique on matched received signals to reduce false targets caused by such clutter. The clutter map technique manages clutter information based on azimuth and elevation angles, designating areas with clutter as unobserved zones to reduce false targets caused by clutter.

[0005] The clutter map technique detects targets by dividing the detection area into small cells, measuring signals over multiple scans within that area to average their magnitudes, and using this as a threshold. However, because the clutter map forms a map through averaging over a long period, it can increase the load on the system by managing a large amount of raw data, and it has the disadvantage of suppressing ECM attacks and low-velocity targets.

[0006] A clutter map was formed using radar received signals and utilized as a simple threshold for target detection; however, target information that was dropped through this method (target signals smaller than the threshold) is not used.

[0007] In this case, the most significant degradation in radar target detection performance in this method is the loss of information regarding actual targets that were not captured near clutter. Furthermore, when false positives caused by clutter occur, they are sometimes treated as targets simply because a threshold was exceeded. This problem causes errors during the clustering and tracking processes, which ultimately degrades the final radar detection performance. The problem to be solved

[0008] The objective of the present invention is to provide an artificial intelligence-based clutter map generation device and method in a clutter environment. means of solving the problem

[0009] An artificial intelligence-based clutter map generation device in a clutter environment according to one embodiment of the present invention may comprise, in a target detection device using radar, a radar signal input unit that receives a current stage surrounding scan signal by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing unit that provides a previous stage estimation parameter based on the surrounding scan signal of the preceding stage of the obtained surrounding scan signal; a parameter generating unit that estimates a current stage parameter using the current stage surrounding scan signal and the previous stage estimation parameter; and a clutter map generating unit that generates a probability distribution for a specific clutter by repeating the parameter estimation and thereby generates a statistics-based clutter map.

[0010] As an example of an embodiment, the AI-based clutter map generation device in the clutter environment may further include a target signal comparison unit that compares the generated statistics-based clutter map with the input signal to find a target signal in the input signal.

[0011] As an example of an embodiment, the clutter map generation unit can extract representative characteristics of the plurality of cells using some of the plurality of cells based on an artificial intelligence technique including Bayesian estimation.

[0012] As an example of an embodiment, the parameter generation unit can receive the current step estimated parameter as an input of a scan signal around the immediately following step, generate a corresponding statistical distribution parameter, and update it.

[0013] As an example of an embodiment, the clutter map generating unit can tag the probability of cluttering for the input surrounding scan signal.

[0014] As an example of an embodiment, the clutter map generating unit may configure area cells for a clutter map representing characteristics in units of multiple cells of the input surrounding scan signal.

[0015] An artificial intelligence-based clutter map generation method in a clutter environment according to an embodiment of the present invention may comprise, in a target detection method using radar, a radar signal input step of receiving a surrounding scan signal obtained by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing step of providing a previous step estimation parameter based on a previous step surrounding scan signal of the obtained surrounding scan signal; a parameter generation step of estimating a current step parameter using the surrounding scan signal and the previous step estimation parameter; and a clutter map generation step of generating a probability distribution for a specific clutter by repeating the parameter estimation and generating a statistics-based clutter map through this.

[0016] As an example of an embodiment, the AI-based clutter map generation method in the clutter environment may further include a target signal comparison step of comparing the generated statistics-based clutter map with the input signal to find a target signal in the input signal.

[0017] As an example, the clutter map generation step may configure area cells for a clutter map that represents characteristics in units of multiple cells of the input surrounding scan signal.

[0018] As an example, the clutter map generation step may extract representative characteristics of the plurality of cells using some of the plurality of cells based on an artificial intelligence technique including Bayesian estimation.

[0019] As an example of an embodiment, the clutter map generation step can tag the probability of cluttering for the input surrounding scan signal. Effects of the invention

[0020] A base clutter map generation device and method according to one embodiment of the present invention can generate a clutter map regardless of interference information caused by instantaneous ECM jamming, etc.

[0021] In addition, the base clutter map generation device and method according to one embodiment of the present invention can obtain accurate results during clustering and target tracking by tagging the clutter probability for the detected signal.

[0022] In addition, the base clutter map generation device and method according to one embodiment of the present invention can naturally transition when the probability distribution changes due to a change in the target tracking environment.

[0023] In addition, the base clutter map generation device and method according to one embodiment of the present invention can distinguish between a peak signal that occurs once and a clutter signal that exists concurrently.

[0024] In addition, the base clutter map generation device and method according to one embodiment of the present invention do not require storing a large amount of signals to perform probability analysis, making it possible to efficiently analyze the clutter distribution without system load. Brief explanation of the drawing

[0025] FIG. 1 is a functional block diagram of an artificial intelligence-based clutter map generation device in a clutter environment according to an embodiment of the present invention. FIG. 2 is a flowchart of an artificial intelligence-based clutter map generation method in a clutter environment according to an embodiment of the present invention. Figure 3 illustrates the process of estimating the parameters of a statistical distribution through the current sample to generate parameters in the parameter generation unit of Figure 1. Figures 4(A) and (B) illustrate the distribution of values ​​of samples in a specific interval when a radar signal is received. Figure 5 illustrates the cumulative distribution of the magnitude of the radar signal received in specific section 3 of Figure 4. Figures 6(A) to (D) illustrate various statistical distribution forms displayed according to the shape of the received radar signal. Figure 7 illustrates the clutter probability for each of the different received radar signals. Figure 8 illustrates the radar measurement results through false probability tagging by the clutter map generation device of Figure 1. Specific details for implementing the invention

[0026] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. The advantages and features of the present invention, and the methods for achieving them, will become clear by referring to the embodiments described below in detail together with the attached drawings. However, the present invention is not limited to the embodiments disclosed below but can be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims. Throughout the specification, the same reference numerals refer to the same components.

[0027] Unless otherwise defined, all terms used in this specification (including technical and scientific terms) may be used in a meaning that is commonly understood by those skilled in the art to which the present invention pertains. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0028] In this specification, terms such as "first," "second," etc. are used to distinguish one component from another, and the scope of rights shall not be limited by these terms. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0029] In this specification, identification symbols (e.g., a, b, c, etc.) for each step are used for convenience of explanation and do not indicate the order of the steps; the steps may occur differently from the specified order unless the context clearly indicates a specific order. That is, the steps may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

[0030] In this specification, expressions such as “have,” “may have,” “include,” or “may include” indicate the presence of such features (e.g., numerical values, functions, operations, or components such as parts) and do not exclude the presence of additional features.

[0031] Additionally, the term “part” as used in this specification refers to software or hardware components such as field-programmable gate arrays (FPGAs) or ASICs, and the “part” performs certain roles. However, the “part” is not limited to software or hardware. The “part” may be configured to reside in an addressable storage medium or may be configured to run one or more processors. Thus, by example, the “part” includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data structures, and variables. The functions provided within the components and “parts” may be combined into a smaller number of components and “parts” or further separated into additional components and “parts.”

[0032] The clutter map function detects targets by dividing the detection area into small cells, measuring signals over multiple scans within that area to average their magnitudes, and using this as a threshold value. Since the clutter map is formed through averaging over a long period, it can increase the load on the system by managing a large amount of raw data and can suppress ECM attacks and low-velocity targets. To address these issues, the inventors propose a new clutter map generation technique applicable to AI-based enhanced radar target detection and a supplementary statistics-based clutter map formation process, which utilizes a novel clutter map formation method.

[0033] Hereinafter, an artificial intelligence-based clutter map generation device and method in a clutter environment according to an embodiment of the present invention will be described in detail with reference to the related drawings.

[0034] FIG. 1 is a functional block diagram of an artificial intelligence-based clutter map generation device in a clutter environment according to an embodiment of the present invention, and FIG. 2 is a flowchart of an artificial intelligence-based clutter map generation method in a clutter environment according to an embodiment of the present invention.

[0035] Figure 3 illustrates the process of estimating the parameters of a statistical distribution through the current sample to generate parameters in the parameter generation unit of Figure 1, Figures 4 (A) and (B) illustrate the distribution of values ​​of samples in a specific section when a radar signal is received, and Figure 5 illustrates the cumulative distribution of the magnitude of the radar signal received in specific section 3 of Figure 4.

[0036] FIGS. 6(A) to (D) illustrate various statistical distribution forms displayed according to the shape of the received radar signal, FIG. 7 illustrates the clutter probability for each of the received different radar signals, and FIG. 8 illustrates the radar measurement results through false probability tagging by the clutter map generation device of FIG. 1.

[0037] Referring to FIG. 1, an artificial intelligence-based clutter map generation device (100) in a clutter environment according to an embodiment of the present invention comprises: a radar signal input unit (110) that receives a current stage surrounding scan signal by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing unit (120) that provides a previous stage estimation parameter based on the surrounding scan signal of the previous stage of the obtained surrounding scan signal; a parameter generation unit (130) that estimates a current stage parameter using the current stage surrounding scan signal and the previous stage estimation parameter; and a clutter map generation unit (140) that generates a probability distribution for a specific clutter by repeating the parameter estimation and thereby generates a statistics-based clutter map.

[0038] An artificial intelligence-based clutter map generation device (100) according to one embodiment of the present invention may further include a target signal comparison unit (150) that compares the generated statistics-based clutter map with the input signal in order to find a target signal in the input signal in the clutter environment.

[0039] The clutter map generation unit (140) can extract representative characteristics of the plurality of cells using some of the plurality of cells based on an artificial intelligence technique including Bayesian estimation.

[0040] The parameter generation unit (130) can receive the current step estimation parameter as input and generate a corresponding statistical distribution parameter to update it.

[0041] The clutter map generation unit (140) can tag the probability of cluttering for the surrounding scan signal received above.

[0042] The clutter map generation unit (140) can configure area cells for a clutter map that represents characteristics in units of multiple cells of the input surrounding scan signal.

[0043] Referring to FIG. 2, an artificial intelligence-based clutter map generation method in a clutter environment according to an embodiment of the present invention may include: a radar signal input step (S110) for receiving an surrounding scan signal obtained by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing step (S120) for providing a previous stage estimation parameter based on a previous stage surrounding scan signal of the obtained surrounding scan signal; a parameter generation step (S130) for estimating a current stage parameter using the surrounding scan signal and the previous stage estimation parameter; and a clutter map generation step (S140) for generating a probability distribution for a specific clutter by repeating the parameter estimation and thereby generating a statistics-based clutter map.

[0044] An artificial intelligence-based clutter map generation method according to one embodiment of the present invention may further include a target signal comparison step (S150) for comparing the generated statistics-based clutter map with the input signal in order to find a target signal in the input signal.

[0045] The clutter map generation step (S140) can configure area cells for a clutter map that represents characteristics in units of multiple cells of the input surrounding scan signal.

[0046] The clutter map generation step (S140) can extract representative characteristics of the plurality of cells using some of the plurality of cells based on an artificial intelligence technique including Bayesian estimation.

[0047] The parameter generation step (S130) can receive the current step estimated parameter as input and generate a corresponding statistical distribution parameter to update it.

[0048] Here, the above statistical parameters are set to a Gamma distribution, and then shape parameters (a,b) are extracted through the Likelihood.

[0049] The clutter map generation step (S140) makes it possible to tag the probability of cluttering for the surrounding scan signal received above, and since the distribution itself becomes the clutter map when a gamma distribution is generated through the estimated a and b, the shape is formed by continuously updating parameters.

[0050] As shown in Fig. 2, the statistical-based clutter map can be configured using accumulated information regarding the magnitude of the received signal.

[0051] In summary, when a single scan signal is input according to the rotation of the radar (S110), parameters are generated, and when the signal for the next scan is input, parameters of a new statistical distribution are generated and updated by referring to the previous parameters (S120, S130). By repeating the above process, a probability distribution representing a specific clutter can be generated, and a statistics-based clutter map is created through this (S140).

[0052] With reference to FIGS. 1 to 8, an artificial intelligence-based clutter map generation apparatus and method in a clutter environment according to an embodiment of the present invention will be described in more detail.

[0053] The clutter map function, also known as Time Averaging CFAR, is a function that cuts a detection area into small cells, measures signals over multiple scans within that area to average their magnitudes, and uses this as a threshold value to enable target detection.

[0054] Over a long period, each cell holds the average size of the clutter generated at that location and uses this information to form a clutter map; however, this method has disadvantages in terms of memory, resolution, ECM attacks, and suppressing low-velocity targets.

[0055] In one embodiment of the present invention, to resolve this, instead of cells at the A / D sample unit level, multiple cells (specific interval cells) are grouped together to form area cells for a clutter map that represents the characteristics, and information is measured for each scan. At this time, representative characteristics of the area must be extracted using only some samples. This can be obtained through artificial intelligence techniques such as Bayesian estimation.

[0056] Here, the cell of the above A / D sample unit (distance interval unit) refers to a single cell (distance cell), and the A / D sampling interval may be determined differently depending on radar performance, purpose, and design philosophy.

[0057] In addition, the size of the specific section cell (the plurality of cell units) can be determined by analyzing the signal during radar operation, allowing the location and size of the specific section to be determined according to the user's selection, such as an area of ​​interest, a specific area, or an area with a lot of clutter.

[0058] Referring to Figure 3, the estimation method using artificial intelligence techniques estimates the parameters of the statistical distribution estimated through the current sample, and when the next sample is input, it creates the entire distribution by updating the previously estimated parameters.

[0059] At this time, representative characteristics of the region must be extracted using only some samples, which can be obtained through artificial intelligence-based estimation techniques.

[0060] To resolve the problems of conventional methods, this approach measures information per scan by grouping multiple cells (specific interval cells) together to form area cells on a clutter map that represents their characteristics, rather than using cells at the A / D sample level.

[0062] The estimation of parameters of a statistical distribution using artificial intelligence techniques follows Equations 1 and 2 below.

[0063]

[0064] Here, = prior distribution, = posterior distribution, = likelihood, = data, = is after observing data.

[0065]

[0066] Here, am.

[0068] Referring to FIGS. 4 to 6, when the radar signal illustrated in FIGS. 4 (A) and (B) is received, the distribution of values ​​of samples in a specific section (3) is plotted as in FIG. 5 and is expressed as a cumulative distribution with respect to the reception size. As the shape of the expressed statistical distribution varies depending on the data, as shown in FIG. 6, different parameters are calculated. In this case, for the samples in the specific section, the range is determined according to user selection.

[0069] By generating a statistical clutter map through AI-based estimation and continuously updating information from multiple scans, it converges into a specific distribution shape. Through the generated clutter map, incoming target signals can be measured with greater accuracy.

[0070] As illustrated in FIGS. 4(A) and (B), accumulating cell information of a certain section of a radar signal results in the data shown in FIGS. 6(A) to (D). By utilizing a statistics-based clutter map, a clutter map can be generated that ignores information caused by instantaneous ECM jamming. Additionally, the probability of a detected signal being clutter can be tagged, and using this information can yield more accurate results during clustering and target tracking.

[0071] Furthermore, it can naturally transition when the probability distribution changes due to environmental variations, such as by distinguishing between one-time peak signals and constantly existing clutter signals. These advantages ultimately improve the radar's target detection performance.

[0072] As shown in Fig. 7, when a clutter map is generated and a target signal (x-axis) is received, the probability of it being clutter is calculated through the clutter probability (y-axis) value at that time.

[0073] Figure 7 illustrates the clutter probability for each of the received different radar signals (S1, S2), and it is possible to indicate that the probability of S1 being clutter is about 40% and the probability of S2 being clutter is about 5%. Alternatively, if a target signal of about -52.5 dBm is input, the probability of clutter is about 10%.

[0074] Referring to FIG. 8, radar measurement results through false probability tagging by a clutter map generating device are shown, displaying the clutter probability for the hit position and showing the actual position and clutter position of the target.

[0075] The base clutter map generation device and method according to one embodiment of the present invention described above can generate a clutter map independently of interference information caused by momentarily occurring ECM jamming, etc., by using a portion of cells of a plurality of cells with an artificial intelligence technique to extract representative characteristics of a plurality of cells and tagging the probability of clutter occurring for a radar input signal, and it is possible to distinguish between a one-time peak signal and a constantly existing clutter signal.

[0076] In addition, since there is no need to store a large amount of signals to perform probability analysis, it is possible to efficiently analyze clutter distribution without system load. Explanation of the symbols

[0077] 110: Radar signal input section 120: Parameter provider 130: Parameter generation section 140: Clutter Map Generation Section 150: Target signal comparator

Claims

Claim 1 An AI-based clutter map generation device in a clutter environment, comprising: a radar signal input unit that receives a current stage surrounding scan signal by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing unit that provides a previous stage estimation parameter based on the surrounding scan signal of the preceding stage of the current stage surrounding scan signal; a parameter generation unit that estimates a current stage parameter using the current stage surrounding scan signal and the previous stage estimation parameter, receives a next stage surrounding scan signal after estimating the current stage parameter, sets a statistical distribution for the received signal magnitude corresponding to a specific clutter corresponding to the next stage surrounding scan signal as a Gamma distribution, extracts shape parameters (a, b) through likelihood, generates a next stage statistical distribution parameter, updates the next stage statistical distribution parameter, and provides it as an estimation parameter for estimating a subsequent stage parameter; and a clutter map generation unit that generates a probability distribution for the specific clutter based on the statistical distribution parameter generated and updated step by step by the parameter generation unit, and generates a statistics-based clutter map using the probability distribution. Claim 2 An artificial intelligence-based clutter map generation device in a clutter environment according to claim 1, further comprising a target signal comparison unit that compares the input signal with the generated statistics-based clutter map to find a target signal in the input signal. Claim 3 In claim 1 or 2, the clutter map generating unit is an artificial intelligence-based clutter map generating device in a clutter environment that forms area cells for a clutter map representing characteristics in multiple cell units of the input surrounding scan signal. Claim 4 In paragraph 3, the clutter map generation unit is an AI-based clutter map generation device in a clutter environment that extracts representative characteristics of the plurality of cells using a portion of the plurality of cells based on an AI technique including Bayesian estimation. Claim 5 delete Claim 6 In claim 1 or 2, the clutter map generating unit is an artificial intelligence-based clutter map generating device in a clutter environment that tags the probability of cluttering for the input surrounding scan signal. Claim 7 A method for generating an AI-based clutter map in a clutter environment, comprising: a radar signal input step of receiving a current stage surrounding scan signal by scanning the surroundings of the radar according to the rotation of the radar; a parameter providing step of providing a previous stage estimation parameter based on the surrounding scan signal of the stage immediately preceding the current stage surrounding scan signal; a step of estimating a current stage parameter using the current stage surrounding scan signal and the previous stage estimation parameter; a step of receiving a next stage surrounding scan signal after estimating the current stage parameter; a step of generating a next stage statistical distribution parameter by setting a statistical distribution for the received signal magnitude corresponding to a specific clutter corresponding to the next stage surrounding scan signal as a Gamma distribution and extracting shape parameters (a, b) through likelihood; a step of updating the next stage statistical distribution parameter and providing it as an estimation parameter for estimating a subsequent stage parameter; and a clutter map generation step of generating a probability distribution for the specific clutter based on the statistical distribution parameter generated and updated step by step, and generating a statistics-based clutter map using the probability distribution. Claim 8 A method for generating an artificial intelligence-based clutter map in a clutter environment according to claim 7, further comprising a target signal comparison step of comparing the input signal with the generated statistics-based clutter map to find a target signal in the input signal. Claim 9 In claim 7 or 8, the clutter map generation step is an artificial intelligence-based clutter map generation method in a clutter environment that comprises constituting area cells for a clutter map representing characteristics in multiple cell units of the input surrounding scan signal. Claim 10 In claim 9, the clutter map generation step is an AI-based clutter map generation method in a clutter environment that extracts representative characteristics of the plurality of cells using a portion of the plurality of cells based on an AI technique including Bayesian estimation. Claim 11 delete Claim 12 In claim 7 or 8, the clutter map generation step is an artificial intelligence-based clutter map generation method in a clutter environment that tags the probability of cluttering for the input surrounding scan signal.

Images