Sonar Data Processing Research Articles

An analysis of reverberation fluctuation by mean square derivation

A method for evaluating the fluctuation of the reverberation envelope is proposed and examined through simulation and real data in this paper. This method is different from the coefficient of variation which is only a function of the first and second order moment of the reverberation statistical model. By using the standard variance of the mean square derivate (MSD) of the reverberation envelope, the paper shows that the reverberation fluctuation is a function of the bandwidth of the emitted signal, and that a large value of the square variance means little fluctuation of the reverberation envelope. Theoretical studies show that the standard variance is proportional to the bandwidth of the emitted signal, so we conclude that less time width or larger bandwidth of the transmitted signal produces less fluctuation. Simulation and real active sonar data processing are used to verify this conclusion.

Higher-Order Statistics for the Detection of Small Objects in a Noisy Background Application on Sonar Imaging

An original algorithm for the detection of small objects in a noisy background is proposed. Its application to underwater objects detection by sonar imaging is addressed. This new method is based on the use of higher-order statistics (HOS) that are locally estimated on the images. The proposed algorithm is divided into two steps. In a first step, HOS (skewness and kurtosis) are estimated locally using a square sliding computation window. Small deterministic objects have different statistical properties from the background they are thus highlighted. The influence of the signal-to-noise ratio (SNR) on the results is studied in the case of Gaussian noise. Mathematical expressions of the estimators and of the expected performances are derived and are experimentally confirmed. In a second step, the results are focused by a matched filter using a theoretical model. This enables the precise localization of the regions of interest. The proposed method generalizes to other statistical distributions and we derive the theoretical expressions of the HOS estimators in the case of a Weibull distribution (both when only noise is present or when a small deterministic object is present within the filtering window). This enables the application of the proposed technique to the processing of synthetic aperture sonar data containing underwater mines whose echoes have to be detected and located. Results on real data sets are presented and quantitatively evaluated using receiver operating characteristic (ROC) curves.

Occupancy grids building by sonar and mobile robot

In this paper, a modified method for occupancy grid map building by a moving mobile robot and a scanning ultrasonic range-finder is proposed. The map building process consists of two phases: (1) gleaning of information from environment, and (2) sonar data processing. For sonar data processing the proposed modified method combines: (1) statistical approach for probability sonar model building; and (2) application of fuzzy logic theory for sonar data fusion. It is experimentally shown that, in some applications, the proposed modified method has advantages over other well-known methods.

Environment map building using mobile robot and ultrasonic range finder

In this paper, a modified method for occupancy grid map building by a moving mobile robot, equipped with a scanning ultrasonic range finder, is proposed. The map building process consists of twophases - gleaning of information from environment and sonar data processing. For sonar data processing the proposed modified method combines the statistical approach for probability sonar model building with the application of fuzzy logic for sonar data fusion. Experimental results are presented.

Echoview as a multibeam sonar data processing and analysis toolkit for fisheries research

Echoview is a hydroacoustic data analysis software package, widely used in the fisheries research and stock assessment communities. Originally developed to handle a variety of single-beam sonar data formats, Echoview has been extended to support multibeam data. Multibeam data logging, lossless compression, and real time beamforming and display are some of the software’s core features. Multibeam data has an additional dimension compared to single-beam data, and a 3D data viewer has been developed providing 3D visualizations of the seabed and fish schools detected by built-in algorithms. Since the multibeam module is just one of many software modules of the Echoview package, data from other sources such as single-beam sonar systems and current profilers can be combined and analyzed together with the multibeam data. The combination of coincident fish density estimates from calibrated single-beam backscatter data with school volume estimates from multibeam data will represent a significant improvement in stock assessment methods. Ongoing research and development will make it possible for Echoview to follow and even set new trends in multibeam water-column data analysis for fisheries research. Features under development include calibration, vessel motion compensation, improved feature detection, and enhanced and animated 3D displays.

Processing and analysis of Simrad multibeam sonar data

The common approach to analysing data collected with multibeam and sidescan sonars is to visually interpret charts of contoured bathymetry and mosaics of seabed images. However, some of the information content is lost by processing the data into charts because this involves some averaging; the analysis might uncover more information if done on the data at an earlier stage in the processing. Motivated by this potential, I have created a software system which can be used to analyse data collected with Simrad EM1000 (shallow water) and EM12 (deep water) multibeam sonars, as well as to generate bathymetry contour charts and backscatter mosaics. The system includes data preprocessing, such as navigation filtering, depth filtering (removal of outlying values), and amplitude mapping using the multibeam bathymetry to correctly position image pixels across the swath. The data attributes that can be analysed include the orientation and slope of the seafloor, and the mean signal strength for each sounding. To determine bathymetry attributes such as slope, the soundings across a number of beams and across a series of pings are grouped and a least-squares plane fitted to them. Bathymetric curvature is obtained by detrending the grouped data using the least-squares plane and fitting a paraboloid to the residuals. The magnitudes and signs of the paraboloid's coefficients reveal depressions and hills and their orientations. Furthermore, the seafloor geology can be classified using a simple combination of these attributes. For example, flat-lying sediments can be classified where the backscatter, slope and curvature fall below specified values.

Environment representation using multiple abstraction levels

This paper describes an approach to building a hierarchy of maplike descriptions of an initially unknown environment using a mobile robot. High-level tasks require a symbolic and qualitative description of space, yet this remains a stable and consistent method for defining specific regions of interest. The paper describes a hierarchy of spatial abstractions and a classification scheme for spatial regions to be used in navigation and mapping. In particular, the problem of maintaining a stable description of the environment that can be used for symbol extraction receives attention. By combining signal-domain analysis for initial processing of sonar and video data in conjunction with geometric processing for symbolic map data constructed from the sonar data, robust navigation becomes possible. Geometric map construction is based on range data, with sonar data used as an example. Recalibration of the robot's position, however, is critical to this task. Two different and complementary forms, of localization algorithm are described. One of these is based on simple geometric modeling of range data points, followed by data-to-object matching. The second method depends less strongly on a simple environment geometry and is based on image data.

Iterative Restoration Algorithm for Real-Time Processing of Broadband Synthetic Aperture Sonar Data

Broadband synthetic aperture sonar (SAS) is a high resolution underwater imaging technique, which uses a digital matched filter for pulse compression followed by aperture synthesis, also referred to as azimuthal matched filtering, to improve resolution. Thus, the processing scheme is equivalent to a two-dimensional matched filter operation, in which the point spread function (PSF) for the particular SAS-geometry considered is correlated with the observation. It can be shown that this processing scheme is suboptimal, because it causes a blurring of the processed image. Therefore, the purpose of this paper is to develop a computationally efficient iterative algorithm for reconstruction, which compares effectively to the matched filter in the processing time, but shows significant improvement in image detail. The proposed iterative restoration algorithm is derived by modifying the optimal gradient method through an adaptive relaxation technique. The adaptivity is introduced to incorporate properties of the true restoration error. This makes the successive approximations approach the exact solution much faster, enabling a visually good convergence within a few iterates. Asymptotic convergence of the proposed iterative algorithm is established. For the experimental results which are shown, the new algorithm performs better on accuracy and compares well on computation time. Application to underwater object imaging using simulated data shows clear improvements compared to a matched filter processing technique.

A deconvolution algorithm for broadband synthetic aperture data processing

Conventional processing of synthetic aperture sonar (SAS) data is equivalent to a two-dimensional matched filter operation. In principle, two-dimensional deconvolution improves the resolution of the processed image. However, its direct implementation is generally impractical, due to numerical problems. The paper discusses the development of iterative algorithms that efficiently perform the deconvolution of broadband synthetic aperture data and gives examples of their application. It is concluded that, in many cases, the proposed approach is preferable to more classical solutions. >

Applying Vector Processing to Active Sonars for Increased Accuracy and an Expanded Operating Envelope

ABSTRACTAdvanced computing technology development now allows considerable computational power to be packaged in small, modular components for low costs. In particular, vector and array processor add‐on boards offer a way to incorporate modern parallel capabilities into existing hardware systems. In addition to cost efficiency, the fact that these boards may be able to plug‐in to existing computer systems provides both design flexibility and a more rapid way of upgrading deployed computer hardware. To take advantage of such processing capabilities, however, algorithm analysis must be performed to determine what functions can be efficiently off‐loaded to the vector processing board, and to decide what parts of the original serial algorithm should be redesigned to take advantage of the new hardware architecture.This paper presents a case study of how these issues are being addressed in a real‐time active sonar data processing system. Vector processing technologies are being analyzed for application to correlation sonar systems used for estimating ship velocity on submarines, surface ships, and autonomous underwater vehicles. Such a system is deployed on Trident II SSBNs and is meeting its current design objectives. However, real‐time operation of the correlation sonar is computationally intensive and fully utilizes the processing power of the fielded computing hardware. Expansions in the accuracy and operating envelope of the system are desirable and are not limited by the physics or practical operation of the sonar system, but rather by the speed at which processing of hydrophone returns can be performed. This analytic feasibility study investigates the possibility of expanding the operating envelope through the use of vector processing technology, within the constraint that any hardware or software changes would have minimal impact on the deployed system. It is predicted that significant speedups and operating envelope expansions can be achieved by introducing vectorized hardware and algorithms into the existing scalar system.

Qualitative and Quantitative Processing of Side-Scan Sonar Data: ABSTRACT

Qualitative and Quantitative Processing of Side-Scan Sonar Data: ABSTRACT

The resolution of overlapping echos

We present a new method for estimating the number and arrival times for overlapping signals with a priori known shape from noisy observations received by a sensor. The method is based on a recently developed eigenstructure technique for multitarget direction finding with passive antenna arrays and exploits the structure of the received signal covariance matrix. This problem is important in various applications such as radar and sonar data processing, geophysical/seismic exploration, and biomedical engineering. In many of these applications, a known signal is launched into a scattering medium and the returning response-in the form of delayed overlapping echos in noise-has to be processed to yield information on the nature and location of scatterers. The method presented also solves more general problems of signal detection and resolution.

Side-scan sonar mapping and computer-aided interpretation in the Santa Barbara Channel, California

We have experimented with digital processing of side scan sonar data taken in a 14 sq-km area of continental shelf offshore Southern California. The data were FM tape recorded during the survey and digitized and processed later in the laboratory. The digital image processing included both image correction and image enhancement. Geometric corrections were applied to correct for image distortions due to variable ship position and speed and sonar slant range. Enhancements that were tried included contrast stretching, band-pass filtering, image restoration (inverse filtering), and various edge enhancements such as density slicing and standard deviation filters. Interpretive procedures were also attempted and included digital mosaicking, stereoscopic viewing, and falsecolor display. The most effective processing was geometric correction combined with contrast stretching. Mosaicking proved difficult due to imprecise navigation (±50 m), but was very effective in increasing the understanding of the geologic structure in the survey area.

Fifth generation digital sonar signal processing

This paper discusses the evolutionary development, which has taken place over the last decade, in digital sonar systems architecture with the application of first, second, and third computers as system controllers for sonar systems. It is the opinion of the authors that, with the arrival of microprocessors, the system controller tasks in real time digital sonars will diminish. We present, as the fourth generation, the present systems which still have a relatively large CPU, assisted by an array of microprocessors under their control for several subtasks which can be handled, more efficiently, locally in the systems. The fifth generation concept is postulated as a further development of this concept. A distributed processing scheme is presented in which the processing elements are actually highly functionally distributed themselves at the lowest level of architecture; consequently, the user views them as uniprocessors within the tightly coupled network. This approach should result in relatively high throughput utilizing a fairly small repertoire of modular hardware components and requiring minimal software effort by implementing, via firmware, very high level macros. This concept allows adaptive system architecture for the various advanced sonar data processing requirements for multielement linear, spatial, or blanket type array systems postulated for the future.

A simple adaptive algorithm for real-time processing in antenna arrays

A new adaptation algorithm designed for real-time data processing in large antenna arrays is presented. The algorithm is used to determine the set of filter coefficients (weights) which minimizes the mean-square error in a multidimensional linear filter. The algorithm forms an estimate of the target signal, which is assumed to be of interest, in the presence of interfering noises. It is assumed that the direction of arrival and spectral density of the target signal are known a priori. No such information is assumed to be available regarding the structure of the interfering noise field. The a priori target information is incorporated directly into the adaptation procedure using a modified gradient descent technique. The mathematical convergence properties of the algorithm are presented and a computer simulation experiment is used as an illustration. It is shown that as the number of iterations becomes large, the expected value of the adaptive solution converges to the minimum mean-square-error solution. It is further shown that the variance of the adapted filter about the optimum solution can be made arbitrarily small by appropriate choice of a scalar constant in the algorithm. These results are based on the assumption that the array signals are Gaussian and that successive time samples are statistically uncorrelated. Thus, the new algorithm is shown to converge to the optimum processor in the limit as the number of adaptations becomes large. Any disadvantage which may arise in the use of such an asymptotically optimum system is offset by the extreme simplicity of the adaptive procedure. This simplicity should prove to be particularly useful in many of the practical array processing problems recently encountered in seismic and sonar data processing.