External Navigation Research Articles

An Ultrasound-Guided Miniature Hand-Eye Integrated Robot for Percutaneous Needle Placement

Abstract Ultrasound-guided percutaneous puncture technology has advantages such as intra-operative real-time imaging, noninvasive operation, high targeting accuracy, nonionizing radiation, and low cost. However, traditional percutaneous puncture surgery requires doctors to hold ultrasound probes or puncture needles, which causes complex operations. In this article, by integrating ultrasound and needle insertion mechanisms, an ultrasound-guided miniature puncture robot is proposed. This robot can work without an external navigation or industrial or cooperative manipulator after hand-eye calibration and acknowledge the coordinate relationship between the ultrasound image and the robot tip. A three degrees-of-freedom (DoFs) of in-plane mechanism using linear actuators is designed so that the puncture needle can always be scanned by ultrasound during the operation, ensuring real-time monitoring. The method of planning puncture path, modeling the robot kinematics, and ultrasound hand-eye calibration are proposed. To track a needle in an ultrasound image, the image recognition and filter algorithm for the needle tip are presented. The accuracy of the robot puncture operation is verified by point targeting and path tracking experiments in the water tank and phantom, and the puncture error of the robot is 1.5 ± 0.5 mm in the water tank, 1.98 mm in the abdominal phantom, and 1.41 mm in the breast phantom. Finally, the work of this article effectively improves the availability and effectiveness of the ultrasound-guided puncture robot.

“Nonbinary can mean different things to different people”: Nonbinary individuals’ internal gender conceptualizations and external navigation and negotiation of gender binarism.

“Nonbinary can mean different things to different people”: Nonbinary individuals’ internal gender conceptualizations and external navigation and negotiation of gender binarism.

Research on Navigation Risk Assessment of Unmanned Ship Under Complex Navigation Conditions

With the research on unmanned ships conducted by the International Maritime Organization, the European Union and China, the navigation risk of unmanned ships has become a hot topic around the world. Based on the research and development history and current situation of unmanned ships at home and abroad, focusing on the three key elements of “Unmanned Ship, shore-based control personnel, external navigation environment”, this study establishes a navigation risk index system for unmanned ships under complex navigation conditions, considering the particularity of the unmanned ship, its perception ability, environmental understanding, situation judgment, communication ability and other indicators. The analytic hierarchy process is used to solve the weight of all kinds of navigation risk factors, and a consistency test of the established index system is carried out. Through expert investigation, the fuzzy membership set of risk indexes is established, and a fuzzy comprehensive evaluation model is established to evaluate the navigation risk of unmanned ships under complex navigation conditions. To avoid the situation of single-factor information being inundated, this study adopts the method of fuzzy comprehensive evaluation from a low level to a high level, and it verifies the algorithm through cases, which proves the validity and rationality of the proposed risk assessment method.

How to Help Older Adults Navigate Through VR? Effects in Outdoor–Indoor Transition Environments

In more complex outdoor–indoor transition scenarios, people were more prone to getting lost due to difficulties in effectively integrating proprioceptive cues and external navigation cues. Previous research had mostly focused on spatial perception driven by head or limb movements, with less discussion on locomotion methods that involved significant visual discrepancies. The challenge in designing navigation systems lies in balancing the enhancement of wayfinding performance with the potential negative impact on spatial cognition. We believe that navigation legibility might be a key optimization approach. Additionally, the differences in performance between different age groups were also worth investigating. Therefore, this study aimed to explore how locomotion methods and navigation legibility affected wayfinding performance, navigation interaction, navigation regression, and spatial cognition in younger and older adults during outdoor–indoor transition scenarios. Twenty-four younger (aged 19–27) and 24 older adults (aged 60–83) underwent VR wayfinding tasks. Participants initially followed virtual phone navigation from outdoor starting points to indoor endpoints, then drew cognitive route maps. Five main findings emerged: Continuous teleportation positively affected navigation performance and global representation of routes. Improving navigation legibility boosted navigation interaction, navigation regression, and global representation of routes. Older adults engaged more in navigation interaction but less in navigation regression, exhibiting poorer performance and route representation. Location-specific differences were observed in locomotion, legibility, and age effects. Furthermore, most older adults and over half of younger adults displayed a deflection effect based on outdoor–indoor (OI) space in route representation. These findings indicated that continuous teleportation and enhanced navigation legibility could improve user experience and spatial cognition, highlighting the importance of considering user age and the characteristics of outdoor–indoor transition scenarios when optimizing navigation methods.

An endoscopic chisel: intraoperative imaging carves 3D anatomical models.

Preoperative imaging plays a pivotal role in sinus surgery where CTs offer patient-specific insights of complex anatomy, enabling real-time intraoperative navigation to complement endoscopy imaging. However, surgery elicits anatomical changes not represented in the preoperative model, generating an inaccurate basis for navigation during surgery progression. We propose a first vision-based approach to update the preoperative 3D anatomical model leveraging intraoperative endoscopic video for navigated sinus surgery where relative camera poses are known. We rely on comparisons of intraoperative monocular depth estimates and preoperative depth renders to identify modified regions. The new depths are integrated in these regions through volumetric fusion in a truncated signed distance function representation to generate an intraoperative 3D model that reflects tissue manipulation RESULTS: We quantitatively evaluate our approach by sequentially updating models for a five-step surgical progression in an ex vivo specimen. We compute the error between correspondences from the updated model and ground-truth intraoperative CT in the region of anatomical modification. The resulting models show a decrease in error during surgical progression as opposed to increasing when no update is employed. Our findings suggest that preoperative 3D anatomical models can be updated using intraoperative endoscopy video in navigated sinus surgery. Future work will investigate improvements to monocular depth estimation as well as removing the need for external navigation systems. The resulting ability to continuously update the patient model may provide surgeons with a more precise understanding of the current anatomical state and paves the way toward a digital twin paradigm for sinus surgery.

Determining the invariant of inter-frame processing for constructing the image similarity metric

The relevance of modeling digital images is determined by the need to implement approaches in the study of localization and identification of objects in order to reduce the amount of data. In this paper, the study object is the topology of a discrete two-dimensional image within the framework of the problem of determining the invariants of diffeomorphic transformations. Geographic information objects (GIOs) refer to objects that are on a given surface or objects that locally change the surface. With regard to objects, it is assumed that the change in their geolocation in the process of forming both single images and an extended series of frames obtained in the process of continuous monitoring is insignificant. In the process of scanning the surface, possible changes in the position of the image source are taken into account, for example, such as yawing, rolling, and pitch in the case of unmanned aerial vehicles (UAVs). These maneuvers are represented as a group of diffeomorphisms that are controlled by the internal gyroscopes of the carrier and the external navigation system. Based on the studies reported here, the initial ontology of digital images (ODI) has been determined by using the model of color spaces and functions of a special kind. The presence of an ontology makes it possible to build an adequate topology of color distribution in the image and take into account the specificity of the distribution of different colors in a digital image. The study results indicate that a promising method is to determine the similarity by constructing a color atlas structure graph (CASG) based on ODI and by determining invariants as a fragment of CASG inherited by all images in the sequence. The scope and conditions for the practical use of the result include its application to the analysis of images by methods of artificial intelligence

СРАВНИТЕЛЬНЫЙ АНАЛИЗ ЦЕНТРАЛИЗОВАННОГО И ДЕЦЕНТРАЛИЗОВАННОГО АЛГОРИТМОВ ДВИЖЕНИЯ СТРОЕМ БЛА МУЛЬТИКОПТЕРНОГО ТИПА

The development of robotics makes their group application relevant for solving varioustasks. The effectiveness of performing the tasks of detecting and determining the coordinates ofobjects by a group of robots significantly depends on the accuracy of maintaining a given formation.In this regard, the task of determining motion planning algorithms that ensure the greatestaccuracy of maintaining a given formation is of practical interest. This article is devoted to thestudy of the accuracy of maintaining the formation of a multicopter-type UAV group using a centralizedmotion planning algorithm and a decentralized algorithm. The centralized algorithm usesa master UAV, which transmits its coordinates to the slave UAVs. Based on the coordinates obtainedand the given framework of the formation, the guided UAVs plan their movement. In a decentralizedsystem, neighboring UAV groups transmit their coordinates to each other, on the basisof which the movement of a separate UAV is planned. The accuracy of the control system is investigateddepending on the errors of the navigation system and the frequency of updating data on theposition of the leading or neighboring UAVs. It is assumed that the group's UAVs determine their coordinates in discrete moments of time using an external navigation system. Centralized anddecentralized algorithms are worked out by the same motion control system. The algorithms areinvestigated in this article by numerical modeling methods. In the process of simulation, models ofkinematics, dynamics and actuators are taken into account, as well as models for the formation oferrors in the navigation system. It is shown that the de-centralized algorithm of group motionplanning provides higher accuracy compared to the centralized algorithm. However, the technicalimplementation of a decentralized algorithm is more complicated from the point of view of organizinga group communication system. In a centralized system, data transmission from the masterUAV to the slave should be implemented. In a decentralized system, it is required to implementnetwork communication.

Use of a neural network to generate a whole atrium 3D reconstruction of electrographic flow and basket catheter geometry from biosignals alone

Abstract Funding Acknowledgements Type of funding sources: None. Background Electrographic flow (EGF) mapping detects and quantifies patterns of electrical wavefront propagation in the atrial myocardium from basket catheter recordings. However, no single catheter can map the entire endocardial surface simultaneously. Creating panoramic maps of the whole atrium may be possible by reconstructing multiple basket positions obtained within a few minutes of each other. Purpose Present a method of 3D reconstruction of EGF maps generated from individual basket recordings using QRS morphologies and CT anatomies. Methods EGF maps were generated from unipolar signals recorded over 1 minute from 64-pole basket catheters. QRS morphologies recorded by each individual electrode were used to train a neural network (NN) on the reconstruction of the basket geometries in 3D. The NN uses the similarities between QRS morphologies to represent multiple recordings in one coherent 3D space and spatially integrate phenomena across multiple recordings without any external navigation system. Results A basket catheter was placed in the superior vena cava of a chronic AF canine model. The catheter was manually pulled back in steps of 5mm under fluoroscopic supervision. After reconstruction of the 3D electrode positions, a successive displacement of 3.5±1.2mm per step was reproduced. We applied the proposed method on recordings of 18 patients where an EGF source was identified. Independently, an expert determined the source locations on a 3rd party navigation system, based on 10 predefined regions in the left atrium. In 83% of the cases, the proposed method projected the sources in the same or adjacent region. Conclusion Novel 3D reconstruction from QRS signatures enables integration of EGF maps from multiple basket recordings that can be aligned with CT anatomy. Due to the real-time EGF map acquisition, synthesizing maps from more than 1 basket position provides a more panoramic visualization of complex atrial flow patterns.

GNSS-IMU-assisted colored ICP for UAV-LiDAR point cloud registration of peach trees

Unmanned aerial vehicle (UAV)-borne light detection and ranging (LiDAR) scanners have been adopted as a promising instrument for plant parameter estimation in agricultural studies recently. However, accurate LiDAR data registration typically requires expensive external navigation devices such as survey-grade global navigation satellite systems (GNSSs) and tactical-grade inertial measurement units (IMUs). Although algorithmic point cloud registration can be an alternative method, the lack of unique landmarks in agricultural fields might bring much difficulty to accurate aerial LiDAR data alignment. In this study, we developed a UAV-LiDAR system employing UAV’s built-in navigation units, and proposed a novel approach for registering UAV-LiDAR data of level agricultural fields utilizing a colored iterative closest point (ICP) algorithm and GNSS location and IMU orientation information from the UAV. The proposed algorithm was tested in a peach tree parameter estimation experiment in comparison to GNSS and IMU-based georeferencing. Using manually measured crown widths in two perpendicular dimensions and heights of 11 trees as evaluation metrics, our proposed algorithm achieved a root mean square error (RMSE) range of 0.05 to 0.2 m depending on the tree parameter and flight altitude, and it was able to register tree point clouds up to 67% more accurately in terms of the extracted tree parameters than the georeferencing method. The results demonstrated the potential of the proposed algorithm being a low-cost solution to crop inspection using single-pass aerial LiDAR point clouds from straight-pathed flights, yet future work is still needed to improve the algorithm’s adaptability to multi-pass LiDAR data of complex landscapes from flights with curved paths.

Glossary of Neurostimulation Terminology: A Collaborative Neuromodulation Foundation, Institute of Neuromodulation, and International Neuromodulation Society Project

ObjectiveConsistent terminology is necessary to facilitate communication, but limited efforts have addressed this need in the neurostimulation community. We set out to provide a useful and updated glossary for our colleagues and prospective patients. Materials and MethodsThis collaborative effort of the Neuromodulation Foundation (NF), the Institute of Neuromodulation (IoN), and the International Neuromodulation Society (INS) expands a glossary first published in 2007 for spinal cord stimulation. Peripheral nerve, dorsal root ganglion, deep brain, and motor cortex stimulation have been added to our scope. Volunteers from the collaborating entities used a nominal group process, consensus development panels, and the Delphi technique to reach consensus on inclusion and definition of terms. We created a glossary suitable for print and for expansion on the websites of the collaborating entities, which will offer the possibility of explaining definitions for a general audience. We excluded proprietary and brand names but included terms that have attracted proprietary interest without becoming brands or trademarks. We made an effort to be inclusive while also being concise and economical with space. ResultsWe identified and defined 91 terms for this print edition and created an accompanying list of acronyms. As appropriate, we provided figures to illustrate the definitions. ConclusionsAlthough we refer to the glossary presented herein as the print edition, it can of course be viewed and searched electronically. NF, IoN, and INS will continue to collaborate on expanded web editions that can include hyperlinks for internal and external navigation. We believe this glossary will benefit our growing field by facilitating communication and mitigating inappropriate use of neurostimulation terms.

Public intent profile towards medical facility visualisation systems in russian federation: a one-stage sociological survey

Background. Medical facility visualisation systems comprise navigation, information panels and visual operation control elements. Visual perception surveys in medical institutions are of great value in medical sociology. Knowing the today’s visitor intent towards visual content inside healthcare facilities will allow a better formulation of most comfortable arrangement of medical services and improvement of healthcare delivery to various visitor categories. Objectives . An assessment of modern public intent towards medical visualisation systems in the Russian Federation. Methods. A total of 7,356 citizens across 85 country subjects contributed to a sociological survey. Data of a formalised multistage stratified territorial random sampling survey were taken as main information source. The Kano model was used to organise intent towards attributes of medical visualisation systems. Depending on two parameters (functionality and satisfaction), the attributes were assigned between the categories of attractive, linear, obligatory, indifferent or undesirable. Results. The highest approval level among the visualisation attributes was found towards accuracy of the information panel content. The high-level of respondent approval were external navigation, compliance to the Russian language style and rules, good-making of the element, content structuring and use of modern information technologies. The lowest approval level was in a visually unified medical institution design. The most contentious respondent intent was observed in relation to colour-coding of visualisation objects. Conclusion. The survey recovered a generally similar public intent profile towards main visualisation system attributes in the Russian Federation. The mandatory included two attributes, “compliance to the Russian language style and rules” and “good-making of visualisation element”. Mandatory attributes are baseline to medical visualisation systems. They do not improve the visitor’s satisfaction, but their absence negatively impacts perception of healthcare visualisation systems.

The Dorsal Lateral Habenula-Interpeduncular Nucleus Pathway Is Essential for Left-Right-Dependent Decision Making in Zebrafish

How animals behave using suitable information to adapt to the environment is not well known. We address this issue by devising an automated system to let zebrafish exploit either internal (choice of left or right turn) or external (choice of cue color) navigation information to achieve operant behavior by reward reinforcement learning. The results of behavioral task with repeated rule shift indicate that zebrafish can learn operant behavior using both internal-directional and external-cued information. The learning time is reduced as rule shifts are repeated, revealing the capacity of zebrafish to adaptively retrieve the suitable rule memory after training. Zebrafish with an impairment in the neural pathway from the lateral subregion of the dorsal habenula to the interpeduncular nucleus, known to be potentiated in the winners of social conflicts, show specific defects in the application of the internal-directional rule, suggesting the dual roles of this pathway.

Mitigating Cardiovascular Risk Through User Informed Clinical Decision Support

As the volume of data within the electronic health record (EHR) increases, there is an evident need for user-friendly and efficient clinical decision support tools developed to assist with patient assessment. Risk calculators, specifically for atherosclerotic cardiovascular disease (ASCVD), are examples of surveillance tools that intend to quantify and predict patient risk of suffering a cardiovascular event. However, despite reported frequent use by clinicians, risk calculators exist largely outside of the EHR, requiring external navigation and increasing the likelihood of user error. Using a mixed methods approach to development, the present research mitigates the challenges posed by external surveillance platforms and discusses the process of designing and optimizing a clinical tool intended to address ASCVD risk at the point of care. These methods ultimately resulted in a risk calculator with both provider- and patient-facing platforms, data autopopulating functionality, and customizable and flexible integration within the provider’s EHR workflow.

A COMPLETE FOS APPROACH FOR INDOOR CROWDSOURCED MAPPING: CASE STUDY ON SAPIENZA UNIVERSITY OF ROME FACULTIES

Abstract. Indoor mapping is an essential process in several applications such as the visualization of space and its utilization, security and resource planning, emergency planning and location-based alerts and, last but not least, indoor navigation. In this work, a completely free and open-source (FOS) approach to map indoor environments, and to navigate through them, is presented. Our tests were carried out within Sapienza University of Rome public buildings; in detail, Letters and Philosophy faculty and Engineering faculty indoor environments were mapped. To reach this goal, only open source software such as Quantum GIS (QGIS) and open-source platforms like Open Street Map (OSM) and its indoor viewer, Open Level Up (OLU) were adopted. A database of indoor environments of the two faculties, completely compatible with OLU, was created through QGIS. In this way, a public territorial information system of classrooms, offices and laboratories is accessible to everyone who can, hence, add or modify the information, following the principle of crowdsourcing and of Volunteered Geographic Information (VGI). The developed procedure is now standard and its outputs accepted by the OSM community. Hence, the long-term developments of this project are the proposal for the volunteered and cooperative indoor mapping and design of strategic buildings and infrastructures (hospitals, schools, public offices, shopping centers, stations, airports etc.), starting from the available information (indoor layouts) and knowledge acquired through experience of people who normally work inside them and/or visit them frequently. In this context it is possible to state that the development of VGI for internal maps for strategic buildings, infrastructures and denied GNSS environments, not only supports and improves internal and external navigation without interruption, but can also have a significant positive impact on security and emergency management.

Path Planner for Autonomous Exploration of Underground Mines by Aerial Vehicles.

This paper presents a path planner solution that makes it possible to autonomously explore underground mines with aerial robots (typically multicopters). In these environments the operations may be limited by many factors like the lack of external navigation signals, the narrow passages and the absence of radio communications. The designed path planner is defined as a simple and highly computationally efficient algorithm that, only relying on a laser imaging detection and ranging (LIDAR) sensor with Simultaneous localization and mapping (SLAM) capability, permits the exploration of a set of single-level mining tunnels. It performs dynamic planning based on exploration vectors, a novel variant of the open sector method with reinforced filtering. The algorithm incorporates global awareness and obstacle avoidance modules. The first one prevents the possibility of getting trapped in a loop, whereas the second one facilitates the navigation along narrow tunnels. The performance of the proposed solution has been tested in different study cases with a Hardware-in-the-loop (HIL) simulator developed for this purpose. In all situations the path planner logic performed as expected and the used routing was optimal. Furthermore, the path efficiency, measured in terms of traveled distance and used time, was high when compared with an ideal reference case. The result is a very fast, real-time, and static memory capable algorithm, which implemented on the proposed architecture presents a feasible solution for the autonomous exploration of underground mines.

Assessment of levels of navigation systems of medical organizations from the position of lean production

Purpose of the study. To evaluate the level of detail, the amount of visualized data and the levels of navigation systems of primary medical organizations.Materials and methods. The object of this study was 33 primary care medical organizations located on the territory of 7 constituent entities of the Russian Federation. The subject of the study was all the internal and external navigation elements of the MO. The assessment was based on the ALIDS checklist (version 1.0) of 72 indicators, united by five characteristics in general characteristics: architectural-planning and design solutions, transmitted information, levels and navigation elements. The “Levels” block was represented by 19 parameters with a separate assessment of the internal level (9 indicators), the intermediate level (5 indicators) and the external level (5 indicators) of navigation. For each criterion, a nominal dichotomous point score was set with the possibility, if necessary (doubt, clarification, question, etc.) of indicating an expert comment.Results. A full-time study of the navigation systems of 33 primary medical organizations found that the levels of intermediate and external visualization are not used enough. Unlike architectural and planning solutions, various levels of navigation are maximally represented in medical institutions of the Tyumen and Kaliningrad regions, Krasnoyarsk Territory. The lack of an external level of navigation contributes to the main losses of lean production: unnecessary movement, unnecessary transportation and expectations from patients and visitors. Among the objects of internal navigation, cabinets and safety signs are most often placed on the navigation elements. The maximum percentage of compliance among the intermediate navigation category has been established for the main entrances for visitors, including entrances for patients with limited mobility.Conclusion. The navigation system of a medical organization is an important part of the visualization system and one of the criteria for the quality of space. The qualimetry of navigation systems not only of medical organizations, but also of any other buildings and premises, is currently not conclusively presented. Using the author's ALIDS method of audit of navigation systems allows us to give a quantitative and qualitative assessment of the navigation elements used in medical institutions. While improving the navigation systems of existing medical organizations and designing/building new ones, especially with complex mixed or section planning, in order to reduce the main losses of lean manufacturing, it is necessary to create the most detailed and understandable general navigation schemes.

Cooperative Relative UAV Attitude Estimation Using DoA and RF Polarization

Relative unmanned aerial vehicle attitude is estimated using only on-board radio-frequency signaling. The method uses a direction-of-arrival (DoA) vector estimate to determine two degrees-of-freedom (DoFs), a polarimetric narrowband multiple-input multiple-output (MIMO) channel estimate to specify the third DoF to within a 180° ambiguity, and one of the several potential methods for ambiguity resolution. Simulation results demonstrate that the method accurately determines aircraft attitude, with errors proportional to DoA and MIMO channel estimate errors. This approach is useful for cooperative navigation when external navigation aids are not available, such as in GPS-denied environments.

Comprehensive calibration algorithm for long-endurance shipborne grid SINS

A comprehensive calibration algorithm is proposed in this paper with the aim of solving the problem whereby the navigation error of a long-endurance shipborne grid strapdown inertial navigation system (SINS) drifts with time. First, the equation within the inertial frame is deduced to establish the relationship between the position error, the grid yaw error and the platform drift angle in the inertial frame; then in combination with the equation within the inertial frame, the calibration schemes are designed. The gyroscope drift in the body frame is estimated and compensated through the designed calibration schemes with the aid of information either on two intermittent external positions and yaw, or on three intermittent external positions. The simulation results illustrate that, in the former case, the three-axis gyroscope drifts can be estimated accurately; in the latter case, the z-axis gyroscope drift and the grid yaw error can be estimated accurately. Resetting the system error based on external navigation information and compensating the gyroscope drift can effectively restrain the accumulated navigation error of shipborne grid SINS; meanwhile the proposed algorithm is unaffected by the motion of the ship, which has significant practical value in engineering.

Brain-Inspired Cognitive Model With Attention for Self-Driving Cars

The perception-driven approach and end-to-end system are two major vision-based frameworks for self-driving cars. However, it is difficult to introduce attention and historical information into the autonomous driving process, which are essential for achieving human-like driving in these two methods. In this paper, we propose a novel model for self-driving cars called the brain-inspired cognitive model with attention. This model comprises three parts: 1) a convolutional neural network for simulating the human visual cortex; 2) a cognitive map to describe the relationships between objects in a complex traffic scene; and 3) a recurrent neural network, which is combined with the real-time updated cognitive map to implement the attention mechanism and long-short term memory. An advantage of our model is that it can accurately solve three tasks simultaneously: 1) detecting the free space and boundaries for the current and adjacent lanes; 2) estimating the distances to obstacles and vehicle attitude; and 3) learning the driving behavior and decision-making process of a human driver. Importantly, the proposed model can accept external navigation instructions during an end-to-end driving process. To evaluate the model, we built a large-scale road-vehicle dataset containing over 40 000 labeled road images captured by three cameras placed on our self-driving car. Moreover, human driving activities and vehicle states were recorded at the same time.

Autonomous Underwater Vehicle Navigation Using Sonar Image Matching based on Convolutional Neural Network

This paper presents an image matching algorithm based on convolutional neural network (CNN) to aid in the navigating of an Autonomous Underwater Vehicle (AUV) where external navigation aids are not available. We aim to solve the problem where traditional image feature representations and similarity learning are not learned jointly and to improve the matching accuracy of sonar images in deep ocean with dynamic backgrounds, low-intensity and high-noise scenes. In our work, the proposed CNN-based model can train the texture features of sonar images without any manually designed feature descriptors, which can jointly optimize the representation of the input data conditioned on the similarity measure being used. The validation studies show the feasibility and veracity of the proposed method for many general and offset cases using collected sonar images.