• 한국컴퓨터정보학회논문지
• /
• 제29권8호
• /
• pp.43-51
• /
• 2024

본 논문은 수중 활동을 주로 하는 다이버를 대상으로 특수 목적용 다이버 마스크를 이용해서 안구 데이터를 획득 및 분석하고, 이를 이용해서 사용자의 시선을 추적하는 방법에 대해 제안한다. 안구 데이터 분석을 위해 자체 제작한 안구 데이터 셋을 구축하였고, YOLOv8-nano 모델을 활용해서 학습 모델을 생성하였다. 학습 모델의 프레임 당 소요 시간은 평균 45.52ms를 달성하였고, 눈을 뜬 상태와 감는 상태를 구별하는 인식 성공률은 99%를 달성하였다. 안구 데이터 분석 결과를 바탕으로 현실 세계 좌표를 매칭할 수 있는 시선 추적 알고리즘을 개발하였다. 이 알고리즘의 검증 결과 x축은 약 1%, y축은 약 6%의 평균 오차율을 나타내는 것을 알 수 있었다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.746-751
• /
• 2003

A method of measuring the length of defects on the wall of the spent nuclear fuel pool using the image processing and a laser slit beam is proposed. Since the defect monitoring camera is suspended by a crane and hinged to the crane hook, the camera viewing direction can not be adjusted to the orientation that is exactly perpendicular to the wall. Thus, the image taken by the camera, which is horizontally rotated along the axis of the camera supporting beam, is distorted and thus, the precise length can not be measured. In this paper, by using the LASER slit beam generator, the horizontally rotated angle of the camera is estimated. Once the angle is obtained, the distorted image can be easily reconstructed to the image normal to the wall. The estimation algorithm adopts a 3-dimensional coordinate transformation of the image plane where both the laser slit beam and the original image of the defects exist. The estimation equation is obtained by using the information of the beam projected on the wall and the parameters of this equation are experimentally obtained. With this algorithm, the original image of the defect taken at arbitrary rotated angle can be reconstructed to an image normal to the wall. From the result of a series of experiments, the accuracy of the defect is measured within 0.6 and 1.3 % error bound of real defect size in the air and underwater, respectively under 30 degree of the inclined angle of the laser slit beam generator. Also, the error increases as the inclined angle increases upto 60 degree. Over this angle, the defect length can not be measured since the defect image disappears. The proposed algorithm enables the accurate measurement of the defect length only by using a single camera and a laser slit beam.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.1060-1065
• /
• 2005

An Independent Administrative Corporation Japan Agency for Marine-Earth Science and Technology (JAMSTEC) is developing light-and-small Autonomous Underwater Vehicles (AUV)$^{1)}$, named 'MR-X1' (Marine Robot Experimental 1), which can cruise, investigate and observe by itself without human's help. In this paper, we consider the motion control problem of 'MR-X1' and derive a controller. Since the dynamic property of 'MR-X1' is changed by the influence of the speed, the mathematical model of 'MR-X1' becomes the nonlinear model. In order to design a controller for 'MR-X1', we generally apply nonlinear control theories or linear control theories with some constant speed situation. If we design a controller by applying Linear Quadratic (LQ) optimal control theory, the obtained controller only compensates t e optimality at the designed speed situation, and does not compensate the stability at another speed situations. This paper proposes a controller design method using Linear Matrix Inequalities (LMIs)$^{2),3),4)}$, which can adapt the speed variation of 'MR-X1'. And examples of numerical analysis using our designed controller are shown.

• 제어로봇시스템학회논문지
• /
• 제19권8호
• /
• pp.661-666
• /
• 2013

A state feedback controller with integration of output error is proposed for way-point tracking of an AUV (Autonomous Underwater Vehicle). For the steering control on the XY plane, the proposed controller uses three state variables (sway velocity, yaw rate, heading angle) and the integral of the steering error, and for the depth control on the XZ plane, it uses four state variables (pitch rate, depth, pitch angle) and the integral of the depth error. From the simulation using Matlab/Simulink, we verify that the performance of the proposed controller is satisfactory within an error range of 1m from the target way-point for arbitrarily chosen sets of consecutive way-points.

• 로봇학회논문지
• /
• 제15권1호
• /
• pp.16-23
• /
• 2020

This paper describes an alignment algorithm that estimates the initial heading angle of AUVs (Autonomous Underwater Vehicle) for starting navigation in a sea area. In the basic dead reckoning system, the initial orientation of the vehicle is very important. In particular, the initial heading value is an essential factor in determining the performance of the entire navigation system. However, the heading angle of AUVs cannot be measured accurately because the DCS (Digital Compass) corrupted by surrounding magnetic field in pointing true north direction of the absolute global coordinate system (not the same to magnetic north direction). Therefore, we constructed an experimental constraint and designed an algorithm based on extended Kalman filter using only inertial navigation sensors and a GPS (Global Positioning System) receiver basically. The value of sensor covariance was selected by comparing the navigation results with the reference data. The proposed filter estimates the initial heading angle of AUVs for navigation in a sea area and reflects sampling characteristics of each sensor. Finally, we verify the performance of the filter through experiments.

• 제어로봇시스템학회논문지
• /
• 제16권8호
• /
• pp.799-803
• /
• 2010

Recently, there is a rising interest on studying bio-inspired robotic fish because of real fish's great maneuverability and high energy efficiency. However, the researches about the robotic fish have not been done so much and there are still lots of problems to use them in the real environment such as in the river. This paper describes a bio-inspired robotic fish 'Ichthus' which is developed in KITECH and has 3 DOF propulsive mechanism. We develop the dynamic motion equation of 'Ichthus' in the underwater environment and analyze response characteristics of 'Ichthus' according to the input parameters of tail fin's amplitude and oscillation frequency. Then we propose control parameters at the various velocities. These parameters are useful to increase energy efficiency and it can be used when the fish robot moves in the real environment, for example, we can propose proper amplitude and oscillation frequency when the fish robot passes through the narrow space between obstacles.

• 제어로봇시스템학회논문지
• /
• 제19권9호
• /
• pp.797-804
• /
• 2013

In this paper, a study of a new hovering six dof underwater robot with redundant horizontal thrusters, titled HAUV (hovering AUV), is presented. The results of study on the structure design, deployment of thrusters, and development of the developed control system of the AUV was presented. For the HAUV structure, a structure design and an analysis of the thrusting system was performed. For navigation, a sensor fusion board which can proceed various sensor signals to identify correct positions and speeds was developed and a total control system including EKF (Extended Kalman Filter) was designed. Rolling, pitching and depth control tests of the HAUV have been performed, and relatively small angle error and depth tracking error results were shown.

• 제어로봇시스템학회논문지
• /
• 제21권10호
• /
• pp.951-957
• /
• 2015

This Paper considers the heading and depth control problem for an underactuated AUV (Autonomous Underwater Vehicle) HW200. The HW200 is a torpedo-type AUV that is developed from Hanwha corporation R&D Center for military operation such as MCM (Mine Counter Measures). The HW200 controls horizontal and vertical motion with two stern plane and two rudder plane. It is well known that fine control of an AUV motion is not easy because of model uncertainties, highly nonlinear and coupled motions. To overcome those kind of uncertainties, a number of control methods have been presented. In this paper, the motion controllers of the HW200 are designed using PD controller design method based on the linear and perturbed model of the typical 6-DOF equations of an AUV, and confirmed the effectiveness of the controller through simulations and field test.

• 로봇학회논문지
• /
• 제12권4호
• /
• pp.385-394
• /
• 2017

Side scanning sonar (SSS) provides valuable information for robot navigation. However using the side scanning sonar images in the navigation was not fully studied. In this paper, we use range data, and side scanning sonar images from UnderWater Simulator (UWSim) and propose measurement models in a feature based simultaneous localization and mapping (SLAM) framework. The range data is obtained by echosounder and sidescanning sonar images from side scan sonar module for UWSim. For the feature, we used the A-KAZE feature for the SSS image matching and adjusting the relative robot pose by SSS bundle adjustment (BA) with Ceres solver. We use BA for the loop closure constraint of pose-graph SLAM. We used the Incremental Smoothing and Mapping (iSAM) to optimize the graph. The optimized trajectory was compared against the dead reckoning (DR).

• 제어로봇시스템학회논문지
• /
• 제18권3호
• /
• pp.285-294
• /
• 2012

Using bearing measurement only, target motion state is not directly obtained so that TMA (Target Motion Analysis) is needed for this situation. TMA is a nonlinear estimation technique used in passive SONAR systems. Also it is the one of important techniques for underwater combat management systems. TMA can be divided to two parts: batch estimation and sequential estimation. It is preferable to use sequential estimation for reducing computational load as well as adaptively to target maneuvers, batch estimation is still required to attain target initial state vector for convergence of sequential estimation. Selection of batch time interval which depends on observability is critical in TMA performance. Batch estimation in general utilizes predetermined batch time interval. In this paper, we propose a new method called the BTIS (Batch Time Interval and Initial State Estimation). The proposed BTIS estimates target initial status and determines the batch time interval sequentially by using a bank of GMM-TS (Gaussian Mixture Measurement-Track Splitting) filters. The performance of the proposal method is verified by a Monte Carlo simulation study.