Scott Reed

An automatic approach to the detection and extraction of mine features in sidescan sonar

Mine detection and classification using high-resolution sidescan sonar is a critical technology for mine counter measures (MCM). As opposed to the majority of techniques which require large training data sets, this paper presents unsupervised models for both the detection and the shadow extraction phases of an automated classification system. The detection phase is carried out using an unsupervised Markov random field (MRF) model where the required model parameters are estimated from the original image. Using a priori spatial information on the physical size and geometric signature of mines in sidescan sonar, a detection-orientated MRF model is developed which directly segments the image into regions of shadow, seabottom-reverberation, and object-highlight. After detection, features are extracted so that the object can be classified. A novel co-operating statistical snake (CSS) model is presented which extracts the highlight and shadow of the object. The CSS model again utilizes available a priori information on the spatial relationship between the highlight and shadow, allowing accurate segmentation of the objects shadow to be achieved.

The fusion of large scale classified side-scan sonar image mosaics

This paper presents a unified framework for the creation of classified maps of the seafloor from sonar imagery. Significant challenges in photometric correction, classification, navigation and registration, and image fusion are addressed. The techniques described are directly applicable to a range of remote sensing problems. Recent advances in side-scan data correction are incorporated to compensate for the sonar beam pattern and motion of the acquisition platform. The corrected images are segmented using pixel-based textural features and standard classifiers. In parallel, the navigation of the sonar device is processed using Kalman filtering techniques. A simultaneous localization and mapping framework is adopted to improve the navigation accuracy and produce georeferenced mosaics of the segmented side-scan data. These are fused within a Markovian framework and two fusion models are presented. The first uses a voting scheme regularized by an isotropic Markov random field and is applicable when the reliability of each information source is unknown. The Markov model is also used to inpaint regions where no final classification decision can be reached using pixel level fusion. The second model formally introduces the reliability of each information source into a probabilistic model. Evaluation of the two models using both synthetic images and real data from a large scale survey shows significant quantitative and qualitative improvement using the fusion approach.

Real time AUV pipeline detection and tracking using side scan sonar and multi-beam echo-sounder

Robust pipeline tracking is critical for AUV technology to succeed in the commercial sector. The paper presents two techniques for reliably detecting and tracking pipelines using multi-beam echo-sounder and side-scan sonar systems. Because of the specific nature of the problem, a lot of prior knowledge can be used. Our algorithms use a model-based Bayesian approach. They are both efficient and robust to variations of the model and noise. Results are shown on real data sets in both cases. The algorithms are compatible with real-time implementation.

Model-based approach to the detection and classification of mines in sidescan sonar

This paper presents a model-based approach to mine detection and classification by use of sidescan sonar. Advances in autonomous underwater vehicle technology have increased the interest in automatic target recognition systems in an effort to automate a process that is currently carried out by a human operator. Current automated systems generally require training and thus produce poor results when the test data set is different from the training set. This has led to research into unsupervised systems, which are able to cope with the large variability in conditions and terrains seen in sidescan imagery. The system presented in this paper first detects possible minelike objects using a Markov random field model, which operates well on noisy images, such as sidescan, and allows a priori information to be included through the use of priors. The highlight and shadow regions of the object are then extracted with a cooperating statistical snake, which assumes these regions are statistically separate from the background. Finally, a classification decision is made using Dempster-Shafer theory, where the extracted features are compared with synthetic realizations generated with a sidescan sonar simulator model. Results for the entire process are shown on real sidescan sonar data. Similarities between the sidescan sonar and synthetic aperture radar (SAR) imaging processes ensure that the approach outlined here could be made applied to SAR image analysis.

A model based approach to mine detection and classification in sidescan sonar

Developments in autonomous underwater vehicle (AUV) technology has shifted the direction of mine-counter-measure (MCM) research towards more automated techniques. This paper presents an automated approach to the detection and classification of mine-like objects using sidescan sonar images. Mine-like objects (MLOs) are first detected using a Markov random field (MRF) model. The highlight and shadow regions of these MLOs are then extracted using a co-operating statistical snake model. Objects which are not identified as false alarms are then considered in a third classification phase. A sonar simulator model considers different possible object shapes, measuring the plausibility of each match. A final classification decision is carried out using Dempster-Shafer theory which allows both monoimage and multiimage classification. Results for all phases are shown on real data.

Reducing false alarms in automated target recognition using local sea-floor characteristics

This paper describes the use of local sea-floor characteristics to train a neural network to remove false alarms from an Automatic Target Recognition (ATR) algorithm. We demonstrate that this reduces the Probability of False Alarm (PFA) in difficult areas without impacting the Probability of Detection (PD) in flat areas. The sea-floor characteristics are calculated from the texture and appearance of clutter on the seafloor. Textural characteristics are extracted using a Dual Tree Wavelet (DTW) transform. Highlight and shadow regions are segmented using Markov Random Field (MRF) and graph cuts. Clutter density and height are calculated from the segmented image. The method is tested by training a neural network to filter the detections from a Haar cascade ATR algorithm. The neural network is trained on the ATR response and the seafloor characteristics. On Synthetic Aperture Sonar (SAS) data we report an average reduction of 50% in the false alarm rate over that of the ATR algorithm. The processing time for an 8000×3000 pixel image is approximately 1 second.

The automatic fusion of classified sidescan sonar mosaics using CML-RTS and Markov random fields

This paper presents a framework for registering and fusing classified sidescan sonar data. It builds on recent advances in navigation and registration for improved mosaicing, applying novel fusion algorithms to integrate data from overlapping sidescan survey lines to produce large scale classified mosaics. While typical mine-counter-measures (MCM) and rapid environmental assessment (REA) missions provide various over-lapping views of the same region of seafloor, research on sidescan image analysis has traditionally concentrated on the analysis of individual images. The available information from the other images, relating to the same region of seafloor, is generally not considered. The image registration and mosaicing process allows this complementary data to be fused, producing an improved final classification result. The sidescan imagery is first pre-processed through the application of advanced radiosity correction algorithms. Following radiosity correction, texture segmentation for the data presented in this paper is achieved using features derived from the averaged normalised power spectral density. The individual classification maps are georeferenced and coregistered using a Concurrent Mapping and Localisation Rauch-Tung-Striebel (CML-RTS) procedure. This uses local landmarks within the individual images and the AUVs navigation data to generate a more accurate and smooth navigation trajectory. This trajectory is used to produce the registered classification mosaics. The coregistered classification results are then fused to produce an improved class mosaic for the entire survey region. The fusion model uses a voting scheme to initialize the seafloor map after which a Markov random field (MRF) model is used to produce the final fused classification mosaic. The entire process (classification, registration and fusion) is demonstrated on real sidescan data taken at the Saclant Centre, La Spezia, Italy.

AUTOTRACKER: Real-Time Pipeline and Cable Tracking Technologies for AUVs

Abstract As the underlying vehicle technologies for AUVs mature, increasing attention is being paid to the development of “smarts”, onboard systems used to enhance and extend the range of autonomous operations that can be undertaken. This paper describes AUTOTRACKER-a modular, distributed architecture for realtime tracking of subsea pipelines and cables. It outlines the sensor and processing techniques used for real-time control of the AUV, together with the details of the highperformance obstacle avoidance system which allows low-altitude, high-speed surveys to be performed.