Suspended underwater object detection system using entropy reduction

Abstract

A suspended underwater object detection system configured for detection of a suspended underwater object in a high-clutter underwater environment configured to be deployed in an unmanned underwater vehicle. The suspended underwater object detection system may be configured to process sonar returns received at each cell of a set of cells to generate entropy estimates for the cells, update the entropy estimates for the cells based on an entropy-reduction search strategy, and when an entropy estimate for any one of the cells falls below a predetermined threshold, identify the one cell as likely containing the suspended underwater object.

Description

SUSPENDED UNDERWATER OBJECT DETECTION SYSTEM USING ENTROPY REDUCTIONCLAIM OF PRIORITY

[0001] This patent application claims the benefit of priority to U.S. Provisional Application Serial No. 63 / 689,346, filed August 30, 2024, which is incorporated by reference herein in its entirety.TECHNICAL FIELD

[0002] Embodiments pertain detection of suspended underwater objects in high-clutter underwater environments.BACKGROUND

[0003] Suspended underwater objects pose extreme danger for ships and submarines. Accurate and efficient detection of suspended underwater objects is therefore of critical importance for navigation.

[0004] One issue with the detection of suspended underwater objects is that a high clutter underwater environment creates unwanted echoes that may mimic a suspended underwater object. Another issue with the detection of suspended underwater objects is the speed of detection: conventional techniques employ time-intensive search strategies making it difficult to detect suspended underwater objects in a timely manner.

[0005] Thus, there are general needs improved techniques for detection of suspended underwater objects.BRIEF DESCRIPTION OF THE DRAWINGS

[0006] FIG. 1 illustrates the operational environment of an unmanned underwater vehicle configured for detecting a suspended underwater object in a high-clutter underwater environment, in accordance with some embodiments.

[0007] FIG. 2 is a functional block diagram of an unmanned underwater vehicle configured for detecting a suspended underwater object in a high-clutter underwater environment, in accordance with some embodiments.

[0008] FIG. 3 is a top view illustrating the operation of the unmanned underwater vehicle of FIG. 1 configured for detecting a suspended underwater object, in accordance with some embodiments.

[0009] FIG. 4 illustrates an entropy calculating and reduction process for detecting a suspended underwater object in a high-clutter underwater environment using an unmanned underwater vehicle, in accordance with some embodiments.DETAILED DESCRIPTION

[0010] The following description and the drawings sufficiently illustrate specific embodiments to enable those skilled in the art to practice them. Other embodiments may incorporate structural, logical, electrical, process, and other changes. Portions and features of some embodiments may be included in, or substituted for, those of other embodiments. Embodiments set forth in the claims encompass all available equivalents of those claims.

[0011] Some embodiments are directed to a suspended underwater object detection system configured for detection of a suspended underwater object in a high-clutter underwater environment. The system is configured to be deployed in an unmanned underwater vehicle. In these embodiments, sonar returns received at each cell of a set of cells are processed to generate entropy estimates for the cells, and the entropy estimates for the cells are updated based on an entropyreduction search strategy. In these embodiments, when an entropy estimate for any one of the cells falls below a predetermined threshold, the cell is identified as likely containing the suspended underwater object.

[0012] Embodiments disclosed herein increases the speed of detecting objects in a high clutter environment. These embodiments balance exploitation and exploration search methods by conducting a search based on reducing entropy. In some embodiments, infotaxis is applied to maximize target-search efficiency. Some embodiments provide for high-speed detection for localizing an object in a low probability of detection environment. These embodiments, as well as others, are described in more detail below.

[0013] FIG. 1 illustrates the operational environment of an unmanned underwater vehicle configured for detecting a suspended underwater object in a high-clutter underwater environment, in accordance with some embodiments.Unmanned underwater vehicle 100 may be configured for detection and destruction of a suspended underwater object 102 in a high-clutter underwater environment 104. The suspended underwater object detection system may be configured to process sonar returns received at each cell of a set of cells to generate entropy estimates for the cells, update the entropy estimates for the cells based on an entropy-reduction search strategy, and when an entropy estimate for any one of the cells falls below a predetermined threshold, identify the one cell as likely containing the suspended underwater object. These embodiments are discussed in more detail below.

[0014] FIG. 2 is a functional block diagram of an unmanned underwater vehicle configured for detecting a suspended underwater object in a high-clutter underwater environment, in accordance with some embodiments. Unmanned underwater vehicle 100 may comprise a system controller, a propulsion system 202, a guidance system 204, a sonar system 206, acoustic communication data link 208, and a suspended underwater object detection system 210. In these embodiments, the suspended underwater object detection system 210 may comprise processing circuitry 212, and memory 214. The processing circuitry 212 may process sonar returns received via the sonar system 206 at each cell of a set of cells to generate entropy estimates for the cells, and update the entropy estimates for the cells based on an entropy -reduction search strategy. In these embodiments, when the entropy estimate for any one of the cells falls below a predetermined threshold, the system 210 may identify the one cell as likely containing a suspended underwater object. The memory may store the entropy estimates.

[0015] In these embodiments, the locations that maximize entropy reduction are selected in a process also balances exploration and exploitation depending on clutter density. In these embodiments, the process may behave more exploratively in high clutter, because the information collected is less likely to be trusted, and more exploitatively in low clutter, because the information collected is more likely to be accurate. This allows for improved suspended underwater object detection in high-clutter environments. A high clutter underwater environment may, for example, be one that creates unwanted echoes that may mimic a suspended underwater object.

[0016] FIG. 3 is a top view illustrating the operation of the unmanned underwater vehicle of FIG. 1 configured for detecting a suspended underwater object, in accordance with some embodiments, detecting a suspended underwater object, in accordance with some embodiments.

[0017] In the embodiments, the suspended underwater object detection system may be configured to determine locations for the cells of the set (i.e., a vector) based on an initial location estimate of the suspended underwater object, the cells being located along an underwater travel path in the horizontal plane at a horizontal distance from the initial location estimate. In these embodiments, the system may also generate signalling to cause the underwater vehicle to move along the underwater travel path to each of the cells and process sonar returns received at each of the cells to generate initial entropy estimates for each of the cells.

[0018] In some embodiments, after an initial entropy estimate is generated for each cell, the system may be configured to update the entropy estimates of the cells by selecting a next one of the cells based on an entropyreduction search strategy, and processing sonar returns for the selected next cell, and calculate an entropy at the selected next cell from the processed sonar returns. In these embodiments, the entropy estimate for all the cells is updated based on the calculated entropy at the selected next cell.

[0019] In some embodiments, after selecting the next one of the cells, the processing circuitry is configured to generate signalling to cause the underwater vehicle to move along the underwater travel path to the selected next cell and configure the underwater vehicle to transmit sonar signals within the selected next cell. In some embodiments, the system may be configured to repeat the updating of the entropy estimates for all the cells until an entropy estimate for one of the cells is below the predetermined threshold, and generate an indication of an actual location of the suspended underwater object based on the sonar returns from the one cell having the entropy estimate below the threshold.

[0020] In these embodiments, the suspended underwater object is assumed to be in the cell when the entropy falls below the predetermined threshold. The threshold may be determined based on what is enough to consider it a detection. These embodiments do not provide a 100% certainty that the suspended underwater object is in a cell, however, when the entropy is lowenough, there is very little likeliness the suspended underwater object could be in another cell.

[0021] In some embodiments, the system may be configured to implement the entropy-reduction search strategy by selecting the next cell with a lowest entropy estimate for transmission of sonar signals and processing the sonar returns, wherein the processed sonar returns from the selected next cell are used to update the entropy estimates for all the cells. In these embodiments, the search strategy balances exploitation and exploration to reduce entropy.

[0022] In some embodiments, the system may be configured to update the entropy estimate for all the cells based on a calculated entropy at the selected next cell using the following equation:

[0023] Mutual Information = Current Entropy-Entropy After Cell B(t+V) is Measured.

[0024] In these embodiments, each cell has an expected amount of information gain called mutual information. The term “Current Entropy” is the entropy of all the cells at the current moment. The term “Entropy After Cell (B+l) is Measured” is the entropy value of all the cells given what cell could be measured next. The term “Mutual Information” is the expected information gain from measuring a cell, B(t+1) and is determined by calculating the expected conditional entropy of each possible cell for B(t+ 1), where ‘B’ is the cell being measured. In these embodiments, after the Mutual Information of each cell is each cell, the processing circuitry may decide which possible B(t+1) should result in the lowest conditional entropy (i.e. the cell that would maximize information gain). The processing circuitry may configure the system to then measures this cell that maximizes information gain and entropy reduction.

[0025] In some embodiments, the Mutual Information may be equal to: IA(t),Bt+l,Xt+l, the Current Entropy may be equal to: HA(t), and the Entropy After Cell B(t+1) is Measured may be equal to H(A(t)\Bt+l,Xt+lf where B(t+1) is the next cell that could be measured. X(t+1) is the measurement of the cell measured next (0 or 1). H(A(t)) is the entropy of all the cells (entropy is a singular value).

[0026] In some embodiments, the system may be configured to update the entropy estimates for the cells based on changes in signal content of thesonar returns, (e.g., the acoustic energy received from the echo location signal received in the frequency range of interest.)

[0027] In some embodiments, the selection of the next one of the cells based on the entropy-reduction search strategy comprises applying an infotaxis process. In these embodiments, the infotaxis process searches by seeking out locations that maximize entropy reduction. The infotaxis process also balances exploration and exploitation depending on clutter density, behaving more exploratively in high clutter, because the information collected is less likely to be trusted, and more exploitatively in low clutter, because the information collected is more likely to be accurate.

[0028] In some embodiments, when a suspended underwater object is detected, the system may determine a distance between the detected suspended underwater object and a sonar array of the unmanned underwater vehicle using the processing sonar returns. In these embodiments, a detection may occur when one cell is identified based on the entropy estimate as likely containing the suspended underwater object.

[0029] FIG. 4 illustrates an entropy calculating and reduction process 400 for detecting a suspended underwater object in a high-clutter underwater environment using an unmanned underwater vehicle, in accordance with some embodiments. The search strategy illustrated in FIG. 4 determines which cell B to measure next. In these embodiments, the search strategy process generates a measurement, X = 1 or X = 0, and passes it to the Bayesian Update function to update the PDF of the search area from p(t) to q(t). The search strategy function then chooses the next cell, B, to measure based on the chosen search strategy, repeating the process with p(t + 1) = q(t). In this diagram, X denotes a measurement. The probability PF A is the probability of a false alarm, while PD is the probability of detection. The variable B(t) is the cell being measured on the tth iteration, while A is the cell where the target is located. The vector p(t) is the current PDF of the search area, while q(t) is the updated PDF of the search area immediately after a measurement has been taken.

[0030] Some embodiments are directed to non-transitory computer- readable storage medium that stores instructions for execution by processing circuitry of a suspended underwater object detection system 210 (FIG. 2)configured for detection of a suspended underwater object 102 (FIG. 1) in a high-clutter underwater environment 104.

[0031] Some embodiments are directed to a suspended underwater object detection system for detection of a suspended underwater object in a high-clutter underwater environment. In these embodiments, the system may process echo location measurements using an initial sonar beam at a first location and apply an infotaxis algorithm to the echo location measurements to select subsequent locations based on a search strategy that reduces entropy. In these embodiments, the system may process subsequent echo location measurements at the subsequent locations and repeat application of the infotaxis algorithm to the echo location measurements from subsequent locations while an entropy associated with a subsequent location exceeds a predetermined threshold. In these embodiments, at each location, the system may analyze the echo location measurements to determine if an entropy associated with the location is at or below the predetermined threshold. In these when the entropy associated with a location is determined to be at or below predetermined threshold, the system may generate signalling indicating a detection of the suspended underwater object. In these embodiments, the processing may continue until a detection occurs or a predetermined time elapses. In these embodiments the vehicle may exit a predefined search area when the entropy associated with a location is determined to be at or below predetermined threshold and the processing stops.

[0032] Some embodiments are directed to an unmanned underwater vehicle 100 (FIG. 1) configured for detection and destruction of a suspended underwater object 102 in a high-clutter underwater environment 104. In these embodiments, the unmanned underwater vehicle 100 may comprise a propulsion system 202 (FIG. 2), a guidance system 204, a sonar system 206, acoustic communication data link 208, and a suspended underwater object detection system 210. In these embodiments, the suspended underwater object detection system 210 may process sonar returns received via the sonar system 206 at each cell of a set of cells to generate entropy estimates for the cells, and update the entropy estimates for the cells based on an entropy-reduction search strategy. In these embodiments, when the entropy estimate for any one of the cells falls below a predetermined threshold, the system 210 may identify the one cell as likely containing a suspended underwater object.

[0033] In some embodiments, the suspended underwater object detection system 210 of the unmanned underwater vehicle may be configured to determine locations for the cells of the set based on an initial location estimate of the suspended underwater object. In these embodiments, the cells may be located along an underwater travel path in the horizontal plane at a horizontal distance from the initial location estimate. The system may generate signalling for the guidance system 204 to cause the underwater vehicle to move along the underwater travel path to each of the cells and process sonar returns received by the sonar system 206 at each of the cells to generate initial entropy estimates for each of the cells.

[0034] In some embodiments, after an initial entropy estimate is generated for each cell, the system may update the entropy estimates of the cells by selecting a next one of the cells based on an entropy-reduction search strategy and processing sonar returns from the sonar system 206 for the selected next cell. The entropy may be calculated at the selected next cell from the processed sonar returns, and the entropy estimate for all the cells may be updated based on the calculated entropy at the selected next cell.

[0035] In some embodiments, after selecting the next one of the cells, the system may generate signalling for the guidance system 204 to cause the underwater vehicle to move along the underwater travel path to the selected next cell and configure the sonar system to transmit sonar signals within the selected next cell. In some embodiments, the system may repeat the updating of the entropy estimates for all the cells until an entropy estimate for one of the cells is below the predetermined threshold and generate an indication for transmission by the acoustic communication data link 208 of an actual location of the suspended underwater object based on the sonar returns from the one cell having the entropy estimate below the threshold.

[0036] In some embodiments, infotaxis is applied as a novel search strategy designed for detecting and locating objects in environments where information is sparse and intermittent, such as the detection of a suspended underwater object in a high-clutter underwater environment. In some embodiments, infotaxis is used to search for suspended underwater object by entropy reduction. In these embodiments, infotaxis treats the search process as an acquisition of information about the source location. Instead of followingconcentration gradients like in chemotaxis, infotaxis uses information as the primary guide. In some embodiments, the process may maintain a probability distribution (Pt(rO)) for the unknown location of the source. This distribution is continuously updated based on the detection events encountered during the search. In these embodiments, the search strategy aims to reduce the entropy of the probability distribution. Entropy quantifies how spread out the distribution is, with lower entropy indicating a more localized source position. In these embodiments, the process chooses a direction that maximizes the expected rate of information acquisition. In these embodiments, the process may balance exploration and exploitation by combining exploitative tendencies (moving towards the most likely source location) with exploratory behavior (gathering more information to improve the estimate). In these embodiments, as the searcher moves and encounters detection events, it updates its probability distribution and recalculates the optimal next move based on the new information.

[0037] Some embodiments, an underwater searcher may use acoustic, magnetic, and / or other sensor data as the primary source of information about potential object locations. In some embodiments, the underwater searcher may maintain and update a probability distribution for the location of objects based on sensor readings and detection events. In some embodiments, the search strategy would aim to reduce the entropy (e.g., the uncertainty) of the probability distribution for object locations in the underwater environment. At each step, the underwater searcher may choose a direction that maximizes the expected rate of information acquisition about potential object locations. In these embodiments, as the underwater searcher moves and encounters detection events, it would update its probability distribution and recalculate the optimal next move based on the new information.

[0038] An example of the use of infotaxis suitable for use is described in “Analysis and Implementation of Infotaxis as a Practical Strategy to Maximize Target-Search Efficiency” by Abigail Rachel Keith, University of Massachusetts Dartmouth Department of Electrical and Computer Engineering, January 2022, which is incorporated herein by reference.

[0039] Embodiments may be implemented in one or a combination of hardware, firmware and software. Embodiments may also be implemented asinstructions stored on a computer-readable storage device, which may be read and executed by at least one processor to perform the operations described herein. A computer-readable storage device may include any non-transitory mechanism for storing information in a form readable by a machine (e.g., a computer). For example, a computer-readable storage device may include readonly memory (ROM), random-access memory (RAM), magnetic disk storage media, optical storage media, flash-memory devices, and other storage devices and media. Some embodiments may include one or more processors and may be configured with instructions stored on a computer-readable storage device.

[0040] The Abstract is provided to allow the reader to ascertain the nature and gist of the technical disclosure. It is submitted with the understanding that it will not be used to limit or interpret the scope or meaning of the claims. The following claims are hereby incorporated into the detailed description, with each claim standing on its own as a separate embodiment.

Claims

CLAIMSWhat is claimed is:

1. A suspended underwater object detection system configured for detection of a suspended underwater object in a high-clutter underwater environment, the system configured to be deployed in an unmanned underwater vehicle, the system comprising: processing circuitry; and memory, the processing circuitry configured to: process sonar returns received at each cell of a set of cells to generate entropy estimates for the cells; update the entropy estimates for the cells based on an entropy-reduction search strategy; and when an entropy estimate for any one of the cells falls below a predetermined threshold, identify the one cell as likely containing the suspended underwater object, wherein the memory is configured to store the entropy estimates.

2. The suspended underwater object detection system of claim 1, wherein the processing circuitry is further configured to: determine locations for the cells of the set based on an initial location estimate of the suspended underwater object, the cells being located along an underwater travel path at a horizontal distance from the initial location estimate; generate signalling to cause the underwater vehicle to move along the underwater travel path to each of the cells; and process sonar returns received at each of the cells to generate initial entropy estimates for each of the cells.

3. The suspended underwater object detection system of claim 2, wherein, after an initial entropy estimate is generated for each cell, the processing circuitry is configured to update the entropy estimates of the cells by: selecting a next one of the cells based on an entropy -reduction search strategy; and processing sonar returns for the selected next cell; andcalculate an entropy at the selected next cell from the processed sonar returns, wherein the entropy estimate for all the cells is updated based on the calculated entropy at the selected next cell.

4. The suspended underwater object detection system of claim 3, wherein after selecting the next one of the cells, the processing circuitry is configured to: generate signalling to cause the underwater vehicle to move along the underwater travel path to the selected next cell; and configure the underwater vehicle to transmit sonar signals within the selected next cell.

5. The suspended underwater object detection system of claim 4, wherein the processing circuitry is further configured to: repeat the updating of the entropy estimates for all the cells until an entropy estimate for one of the cells is below the predetermined threshold; and generate an indication of an actual location of the suspended underwater object based on the sonar returns from the one cell having the entropy estimate below the threshold.

6. The suspended underwater object detection system of claim 5, wherein the processing circuitry is configured to implement the entropy-reduction search strategy by selecting the next cell with a lowest entropy estimate for transmission of sonar signals and processing the sonar returns, wherein the processed sonar returns from the selected next cell are used to update the entropy estimates for all the cells.

7. The suspended underwater object detection system of claim 6, wherein the processing circuitry is configured to update the entropy estimate for all the cells based on a calculated entropy at the selected next cell using equation:Mutual Information = Current Entropy-Entropy After Cell B(t+V) is Measured.

8. The suspended underwater object detection system of claim 7, wherein the processing circuitry is configured to update the entropy estimates for the cells based on changes in signal content of the sonar returns.

9. The suspended underwater object detection system of claim 3, wherein the selection of the next one of the cells based on the entropy -reduction search strategy comprises applying an infotaxis process.

10. The suspended underwater object detection system of claim 3, wherein when a suspended underwater object is detected, the processing circuitry is configured to determine a distance between the detected suspended underwater object and a sonar array of the unmanned underwater vehicle using the processing sonar returns.

11. A non-transitory computer-readable storage medium that stores instructions for execution by processing circuitry of a suspended underwater object detection system configured for detection of a suspended underwater object in a high-clutter underwater environment, the system configured to be deployed in an unmanned underwater vehicle, the processing circuitry configured to: process sonar returns received at each cell of a set of cells to generate entropy estimates for the cells; update the entropy estimates for the cells based on an entropy-reduction search strategy; and when an entropy estimate for any one of the cells falls below a predetermined threshold, identify the one cell as likely containing the suspended underwater object.

12. The non-transitory computer-readable storage medium of claim 11, wherein the processing circuitry is further configured to: determine locations for the cells of the set based on an initial location estimate of the suspended underwater object, the cells being located along an underwater travel path at a horizontal distance from the initial location estimate;generate signalling to cause the underwater vehicle to move along the underwater travel path to each of the cells; and process sonar returns received at each of the cells to generate initial entropy estimates for each of the cells.

13. The non-transitory computer-readable storage medium of claim 12, wherein, after an initial entropy estimate is generated for each cell, the processing circuitry is configured to update the entropy estimates of the cells by: selecting a next one of the cells based on an entropy -reduction search strategy; and processing sonar returns for the selected next cell; and calculate an entropy at the selected next cell from the processed sonar returns, wherein the entropy estimate for all the cells is updated based on the calculated entropy at the selected next cell.

14. The non-transitory computer-readable storage medium of claim 13, wherein after selecting the next one of the cells, the processing circuitry is configured to: generate signalling to cause the underwater vehicle to move along the underwater travel path to the selected next cell; and configure the underwater vehicle to transmit sonar signals within the selected next cell.

15. A suspended underwater object detection system for detection of a suspended underwater object in a high-clutter underwater environment, the system comprising processing circuitry; and memory, the processing circuitry to: process echo location measurements using an initial sonar beam at a first location; apply an infotaxis algorithm to the echo location measurements to select subsequent locations based on a search strategy that reduces entropy; process subsequent echo location measurements at the subsequent locations; andrepeat application of the infotaxis algorithm to the echo location measurements from subsequent locations while an entropy associated with a subsequent location exceeds a predetermined threshold, wherein at each location, the processing circuitry is configured to analyze the echo location measurements to determine if an entropy associated with the location is at or below the predetermined threshold, and wherein when the entropy associated with a location is determined to be at or below predetermined threshold, the processing circuitry is configured to generate signalling indicating a detection of the suspended underwater object.

16. An unmanned underwater vehicle configured for detection of a suspended underwater object in a high-clutter underwater environment, the unmanned underwater vehicle comprising: a propulsion system; a guidance system; a sonar system; acoustic communication data link; and a suspended underwater object detection system comprising processing circuitry configured to: process sonar returns received via the sonar system at each cell of a set of cells to generate entropy estimates for the cells; update the entropy estimates for the cells based on an entropy-reduction search strategy; and when the entropy estimate for any one of the cells falls below a predetermined threshold, identify the one cell as likely containing a suspended underwater object.

17. The unmanned underwater vehicle of claim 16, wherein the processing circuitry is further configured to: determine locations for the cells of the set based on an initial location estimate of the suspended underwater object, the cells being located along an underwater travel path at a horizontal distance from the initial location estimate; generate signalling for the guidance system to cause the underwater vehicle to move along the underwater travel path to each of the cells; andprocess sonar returns received by the sonar system at each of the cells to generate initial entropy estimates for each of the cells.

18. The unmanned underwater vehicle of claim 17, wherein, after an initial entropy estimate is generated for each cell, the processing circuitry is configured to update the entropy estimates of the cells by: selecting a next one of the cells based on an entropy -reduction search strategy; and processing sonar returns from the sonar system for the selected next cell; and calculate an entropy at the selected next cell from the processed sonar returns, wherein the entropy estimate for all the cells is updated based on the calculated entropy at the selected next cell.

19. The unmanned underwater vehicle of claim 18, wherein after selecting the next one of the cells, the processing circuitry is configured to: generate signalling for the guidance system to cause the underwater vehicle to move along the underwater travel path to the selected next cell; and configure the sonar system to transmit sonar signals within the selected next cell.

20. The unmanned underwater vehicle of claim 19, wherein the processing circuitry is further configured to: repeat the updating of the entropy estimates for all the cells until an entropy estimate for one of the cells is below the predetermined threshold; and generate an indication for transmission by the acoustic communication data link of an actual location of the suspended underwater object based on the sonar returns from the one cell having the entropy estimate below the threshold.

Images