• 한국멀티미디어학회논문지
• /
• 제20권8호
• /
• pp.1421-1429
• /
• 2017

Human activity recognition has been studied using various sensors and algorithms. Human activity recognition can be divided into sensor based and vision based on the method. In this paper, we proposed an activity recognition system using acceleration sensor and gyroscope sensor in smartphone among sensor based methods. We used Deep Belief Network (DBN), which is one of the most popular deep learning methods, to improve an accuracy of human activity recognition. DBN uses the entire input set as a common input. However, because of the characteristics of different time window depending on the type of human activity, the RBMs, which is a component of DBN, are configured hierarchically by combining them from different time windows. As a result of applying to real data, The proposed human activity recognition system showed stable precision.

• 제어로봇시스템학회논문지
• /
• 제16권12호
• /
• pp.1201-1207
• /
• 2010

This paper shows to stabilize a balancing robot. We derive the dynamics of a balancing robot and design its controller using LQR method. For stabilizing balancing robot, we introduce a method to detect an angle using inertial sensors. In this study, we use a complementary filter to fuse signals by frequency response of gyroscope and accelerometer in order to measure the inclined angle of balancing robot. The filter coefficients are obtained by least square to minimize error in angle-detecting filter design. And then, after we derive a dynamics of balancing robot using Lagrange method, we linearize that dynamics for using LQR method.

• Journal of Multimedia Information System
• /
• 제5권2호
• /
• pp.115-120
• /
• 2018

Recently, ways for accurately measuring the three dimensional movements of hand are actively researched so as to utilize the measurement data for therapeutic and rehabilitation programs. This research paper aims to introduce a product called Myo Armband, a wearable device comprised of a 3-axis accelerometer, a 3 axis gyroscope, and electromyographic sensors. We compare Armband's performance with that of the Motion Capture System, which is known as a device for providing fairly accurate measurements for angular movements of objects. Dart throwing and wrist winding motions comprised movement scenarios. This paper also discusses one of Armband's advantages - portability, and suggests its potential as a substitute for previously used devices. Decent levels of measurement accuracy were obtained which were comparable to that of three dimensional measurement device.

• 센서학회지
• /
• 제26권1호
• /
• pp.7-14
• /
• 2017

A threshold-based fall recognition algorithm using a tri-axial accelerometer and a bi-axial gyroscope mounted on the skin above the upper sternum was proposed to recognize fall-like activities of daily living (ADL) events. The output signals from the tri-axial accelerometer and bi-axial gyroscope were obtained during eight falls and eleven ADL action sequences. The thresholds of signal vector magnitude (SVM_Acc), angular velocity (${\omega}_{res}$), and angular variation (${\theta}_{res}$) were calculated using MATLAB. When the measured values of SVM_Acc, ${\omega}_{res}$, and ${\theta}_{res}$ were compared to the threshold values (TH1, TH2, and TH3), fall-like ADL events could be distinguished from a fall. When SVM_Acc was larger than 2.5 g (TH1), ${\omega}_{res}$ was larger than 1.75 rad/s (TH2), and ${\theta}_{res}$ was larger than 0.385 rad (TH3), eight falls and eleven ADL action sequences were recognized as falls. When at least one of these three conditions was not satisfied, the action sequences were recognized as ADL. Fall-like ADL events such as jogging and jumping up (or down) have posed a problem in distinguishing ADL events from an actual fall. When the measured values of SVM_Acc, ${\omega}_{res}$, and ${\theta}_{res}$ were applied to the sequential processing algorithm proposed in this study, the sensitivity was determined to be 100% for the eight fall action sequences and the specificity was determined to be 100% for the eleven ADL action sequences.

• 한국통신학회논문지
• /
• 제39C권1호
• /
• pp.82-89
• /
• 2014

최근 M2M/IoT에 대한 관심이 높아지면서 디바이스의 위치와 자세 등을 인식할 수 있는 동작 인식 센서의 필요성이 대두되고 있다. 본 논문에서는 소형의 디바이스에 적합하도록 저잡음, 저전력, 초소형 6축 관성센서 IC를 구현하였다. 본 논문에서 구현된 IC는 3축의 압전형 자이로 센서와 3축의 압저항형 가속도 센서를 사용하며, 3축의 자이로스코프 감지 회로, 자이로스코프 센서 구동 회로, 3축의 가속도 감지 회로, 16bit sigma-delta ADC, 디지털 필터와 제어 회로로 구성되어 있다. 본 IC은 TSMC $0.18{\mu}m$ mixed signal CMOS공정으로 개발되었으며, STM사의 6축 관성 센서인 LSM330의 소비전류 6.1mA보다는 약 27% 낮은 4.5mA의 소비 전류로 동작한다.

• 제어로봇시스템학회논문지
• /
• 제7권5호
• /
• pp.454-464
• /
• 2001

In this paper, a personal navigation system is developed using GPS and dead reckoning sensors. This personal navigation system can be used to track a person inside a building, on an urban street, and in the mountain area. GPS can provide accurate absolute position information, but it cant be used without receiving enough satellite signals. Although the inertial sensors such as gyro an accelerometer and be used without this diggiculty, the inertial sensors severely suffer from their drift errors and the magne-tometer can be easily distorted by surrounding electromagnetic field. GPS and DR sensors can be inte-grated together to overcome these problems. A new personal navigation system that can be carried wit person is developed. A pedometer. actually vertically mounted accelerometer, detects ones footstep and gyro detects heading angle. These DR sensors are integrated with GPS and the humans walking pattern provides additional navigation information for compensating the DR sensors. The field testes are performed to evaluated the proposed navigation algorithm.

• 대한안전경영과학회지
• /
• 제20권2호
• /
• pp.37-44
• /
• 2018

The Location Based Service is growing rapidly nowadays due to the universalization of the use for smartphone, and therefore the location determination technology has been placed in a very important position. This study suggests an algorithm that can provide the estimate of users' location by using smartphone sensors. And in doing so we will propose a methodology for the creation and update of indoor map through the more accurate position estimation using smartphone sensors such as acceleration sensor, gyroscope sensor, geomagnetic sensor and rotation sensor.

• 센서학회지
• /
• 제19권3호
• /
• pp.191-196
• /
• 2010

In this paper, modeling of the non-ideal frequency response, especially "notch-and-spike" magnitude phenomenon and phase lag distortion, are discussed. To characterize the non-ideal frequency response, a new electro-mechanical simulation model based on SPICE is proposed using the driving loop of the capacitive vibratory gyroscope. The parasitic components of the driving loop are found to be the major factors of non-ideal frequency response, and it is verified with the measurement results.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.876-881
• /
• 2003

Reliable attitude estimation during the launch and early orbit phase is a critical issue for a satellite. Typically gyroscopes and other sensors are utilized to estimate the attitude during this phase. It is difficult to estimate the attitude quickly and reliably using gyroscopes because it requires a large computational load and accurate sensor measurements. Furthermore, the gyroscope failure may lead to the loss of the satellite. This paper suggests a simple attitude estimation method of a low earth orbit satellite without using gyroscopes, but only using sun sensors and magnetometers in the sun-acquisition mode. Using Kompsat-I telemetry data, we verified that the suggested algorithm provides attitude estimation within 3 degrees on each axis.

• 제어로봇시스템학회논문지
• /
• 제6권12호
• /
• pp.1120-1125
• /
• 2000

This paper presents a technique for estimating the attitude of a model helicopter at near hovering using a combination of inertial and non-inertial sensors such as gyroscope and potentiometer. To estimate the attitude of helicopter a simplified indirect Kalman filter based on sensor modeling is derived and the characteristics of sensors are studied, which are used in determining the optimal Kalman gain. To verify the effectiveness of the proposed algorithm simulation results are presented with real flight data. Our approach avoids a complex dynamic modeling of helicopter and allows for an elegant combination of various sensor data with different measurement frequencies. We also describe the method of implementation of the algorithm in the model helicopter.