Browsing by Author "Hu, Jwu-Sheng"

Now showing 1 - 20 of 99

3-D human posture recognition system using 2-D shape features
(2007)
Hu, Jwu-Sheng
;
Su, Tzung-Min
;
Lin, Pei-Ching
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents an integrated framework for recognizing 3D human posture from 2D images. A flexible combinational algorithm motivated by the novel view expressed by Cyr and Kimia [1] is proposed to generate the aspects of 3D human postures as the posture prototype using features extracted from the collected 2D images sampled at random intervals from the viewing sphere. Frequency and phase information of the posture are calculated from the Fourier descriptors (FDs) of the sampled points on the posture contour as the main and assistant features to extract the characteristic views as the aspects. Moreover, a modified particle filter is applied to improve the robustness of human posture recognition for continuous monitoring. Experimental trials on synthetic and real sequences have shown the effectiveness of the proposed method.
3D Hand Posture Estimation and Task Semantic Monitoring Technique for Human-Robot Collaboration
(2013-01-01)
Hu, Jhen-Jia
;
Li, Hui-Chieh
;
Wang, Hau-Wei
;
Hu, Jwu-Sheng
;
電控工程研究所
We develop a monitoring system for human-robot collaboration (HRC) of hybrid automation. The system enables the production line to avoid product defects as well as guarantee a real-time warning system in a cellular manufacturing (CM). A task semantic monitoring system includes the three-dimensional (3D) hand posture estimation and the motion trajectory monitoring. The coordinate of hand, the motion of hand, and the semantic of hand are identified by two integrated modules. The verification is executed by applying the system to the assembly lines for power protector.
3D object tracking using mean-shift and similarity-based aspect-graph modeling
(2007)
Hu, Jwu-Sheng
;
Su, Tzung-Min
;
Juan, Chung-Wei
;
Wang, George
;
電控工程研究所
The mean shift algorithm is a popular method in the field of 2D object tracking due to its simplicity and robustness over slight variations of lighting condition, scale and view-point over time. However, the appearance of 3D object might have distinctive variations for different viewpoints over time. In this work, a novel method for tracking 3D objects using mean-shift algorithm and a 3D object database is proposed to achieve a more precise tracking. A 3D obctject database using similarity-based aspect-graph is built from 2D images sampled at random intervals from the viewing sphere. Contour and color features of each 2D image are used for modeling the 3D object database. To conduct tracking, a suitable object model is selected from the database and the mean-shift tracking is applied to find the local minima of a similarity measure between the color histourams of the object model and the target image. The effectiveness of the proposed method is demonstrated by experiments with objects rotating and translating in space.
A Ball-Throwing Robot with Visual Feedback
(2010-01-01)
Hu, Jwu-Sheng
;
Chien, Ming-Chih
;
Chang, Yung-Jung
;
Su, Shyh-Haur
;
Kai, Chen-Yu
This work presents a robot system for throwing a ball into a basket. A stereo vision system is used to measure the position of the target in 3D space. The ball-throwing transformation between the input command of the robot system and the target position is built by cubic polynomial. Through ball-throwing transformation with visual feedback for target position, the robot throws the ball toward the target which can randomly move to everywhere in visible field. The percentage of successful ball-throwing for target within three meters was found to be approximately 99%.
A Discontinuous Space Vector PWM Algorithm in abc Reference Frame for Multilevel Three-Phase Cascaded H-Bridge Voltage Source Inverters
(2017-11-01)
Hu, Jwu-Sheng
;
Lin, Jau-Nan
;
Chen, Hung-Chi
;
電機工程學系
;
Department of Electrical and Computer Engineering
Previous space vector pulse-width modulation (SVPWM) methods for a multilevel three-phase voltage source inverter (VSI) have the problems including the computational complexity to identify the location of the reference voltage vector, the memory requirements of switching states and corresponding duty ratios, or no generalized switching pulse durations of switches. These problems are related to the level number, and they become critical at a high level number. In this paper, a discontinuous SVPWM algorithm in the abc reference frame is first proposed for a multilevel three-phase cascaded H-bridge (CHB) VSI. The L-level three-phase SVPWM problem is solved based on the two-level three-phase SVPWM one. Thus, the computational complexity to determine the nearest three switching states and corresponding duty ratios is independent of the level number. Furthermore, the proposed algorithm provides a systematic approach to calculate the switching pulse durations for a multilevel three-phase CHB VSI, and the approach can be extended to different multilevel three-phase VSIs. The proposed algorithm has no memory requirement and low computational complexity. It is verified by the simulation and experimental results for a five-level three-phase CHB VSI.
A Filtered SVPWM for Multiphase Voltage Source Inverters Considering Finite Pulse-Width Resolution
(2012-07-01)
Chen, Keng-Yuan
;
Hu, Jwu-Sheng
;
電機工程學系
;
Department of Electrical and Computer Engineering
A filtered space-vector pulse-width modulation (SVPWM) considering finite pulse-width resolution is proposed to produce a switching sequence with reduced baseband harmonics for multiphase voltage source inverters (VSI). The conventional SVPWM is sensitive to the pulse-width resolution which leads to increased harmonic distortion in digital-based implementations. By incorporating a pseudofeedback loop regarding weighted voltage difference between desired and produced phase voltages, the quantization error induced by finite pulse-width resolution is compensated, yielding a better tracking performance. The gating signals for multiphase VSIs are presented through rigorous mathematical derivation. In simulations and experiments, harmonic distortions of SVPWMs with and without filter compensation under various pulse-width resolutions are shown. The results confirm the effectiveness of the proposed scheme. The main advantage realized is reduction of acoustic noise and increased tracking accuracy.
A General Actuator Array Amplifier Design Using MIMO Optimal Feedback Quantization
(2008)
Hu, Jwu-Sheng
;
Chen, Keng-Yuan
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper proposes a novel switch-mode power stage architecture for actuator array. When the number of actuator is large, the associated power amplifier becomes costly and bulky if each actuator is controlled by an independent amplifier. The proposed architecture allows sharing of the power amplifier among actuators, and thereby reducing the number of switching devices. The architecture is mapped into the graph theorem and the acceptable actuation states are analyzed. The switching method is controlled by an MIMO optimal feedback quantization, which is designed to minimize a weighted measure of the quantization distortion. Simulation results are provided to illustrate the effectiveness of the proposed scheme.
A General Switch-Mode Amplifier Design for Actuators Using MIMO Optimal Feedback Quantization
(2009-12-01)
Hu, Jwu-Sheng
;
Chen, Keng-Yuan
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper proposes a novel switch-mode power stage architecture for actuators. For a large number of actuators, the associated power amplifier becomes costly and bulky when each actuator is controlled by an independent amplifier. The proposed architecture allows actuators to share power amplifiers, thereby reducing the number of switching devices. This architecture is mapped into the graph theorem, and acceptable actuation states are analyzed. The switching method is controlled by multiple-input multiple-output optimal feedback quantization, which is designed to minimize a weighted measure of quantization distortion. Simulation and experimental results illustrate the effectiveness of the proposed scheme.
A Kinematic Human-Walking Model for the Normal-Gait-Speed Estimation Using Tri-Axial Acceleration Signals at Waist Location
(2013-08-01)
Hu, Jwu-Sheng
;
Sun, Kuan-Chun
;
Cheng, Chi-Yuan
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This study aims at estimating the human walking speed using wearable accelerometers by proposing a novel virtual inverted pendulum model. This model not only keeps the important characteristics of both the biped rolling-foot and the inverted pendulum model, but also makes the speed estimation feasible using human body acceleration. Rather than using statistical methods, the proposed kinematic walking model enables calibration of the parameters during walking using only one tri-axial accelerometer on the waist that collects the user's body acceleration. In addition, this model also includes the effect of rotation of the waist within a walking cycle, which improves the estimation accuracy. Experimental results for a group of humans show a 0.58% absolute error mean and 0.72% error deviation, which is far better than the results of other known studies with accelerometers mounted on the upper body.
A Model-based Human Walking Speed Estimation Using Body Acceleration Data
(2012)
Hu, Jwu-Sheng
;
Sun, Kuan-Chun
;
Cheng, Chi-Yuan
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This study aims at estimating the human walking speed using wearable accelerometers by proposing a novel virtual inverted pendulum model. This model not only keeps the important characteristic in biped rolling-foot model, but also makes the speed estimation feasible using human body acceleration. Rather than the statistical methods, the proposed kinematic walking model enables calibration of the parameters during walking using only one tri-axial accelerometer on waist that reflects the user's inertia information during walking. In addition, this model also includes the effect of rotation of waist within a walking cycle that improves the estimation accuracy. Experimental results on a group of humans show a 1.22% error mean and 2.78% deviation, which is far better than other known studies.
A New Design on Multi-Modal Robotic Focus Attention
(2008)
Lin, Chia-How
;
Yang, Chia-Hsing
;
Wang, Cheng-Kang
;
Song, Kai-Tai
;
Hu, Jwu-Sheng
Human detection and tracking is important for user-friendly human-robot interaction. The robot should be able to find the user autonomously and keep its attention to the user in a human-like manner. In this paper, a design and experimental study of robust human detection and tracking is presented through fusion several modalities of sensory information. The multi-modal interaction design utilizes a combination of visual, audio, and laser scanner data for reliable detection and tracking of an interested user. During tracking motion, obstacle avoidance behavior will be activated any time required to ensure safety. Furthermore, user can further assign the robot to interact with other user by speech command. Experimental results show that the robot can robustly tracks person under complex scenarios.
A new spatial-color mean-shift object tracking algorithm with scale and orientation estimation
(2008)
Juan, Chung-Wei
;
Hu, Jwu-Sheng
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, we propose a new mean-shift tracking algorithm based on a novel similarity measure function. The joint spatial-color feature is used as our basic model elements. The target image is modeled with the kernel density estimation and the new similarity measure functions is developed using the expectation of the estimated kernel density. With these new similarity measure functions, two similarity-based mean-shift tracking algorithms are derived. To enhance the robustness, the weighted background information is added into the proposed tracking algorithm. In order to solve the object deformation problem, the principal component analysis is used to update the orientation of the tracking object, and corresponding eigenvalues are used to monitor the scale of the object. The experimental results show that the new similarity-based tracking algorithms can be implemented in real-time and are able to track the moving object with an automatic update of the orientation and scale.
A Novel Magnetic Dipoles Localization and Mapping Algorithm using Magnetometer Array
(2019-01-01)
Hsieh, Shih-Ho
;
Lo, Sheng-Wen
;
Hu, Jwu-Sheng
;
Chang, Jen-Yuan
;
交大名義發表
This study is about localization and mapping with magnetic landmarks. It aims to find a linear solution to estimate the magnetic dipoles' position and orientation with several magnetometers. With the poses of the magnetic dipoles, localization and mapping can be done. Due to the linear solution, avoiding heavy computation load and a long time to compute is possible. In order to differentiate the magnets from each other, we suppose that the magnetic field is modulated in advance and can be demodulated afterwards. The simulation result shows the effectiveness of the proposed algorithm.
A novel switch mode amplifier design for actuator array using MIMO optimal feedback quantization
(2007)
Hu, Jwu-Sheng
;
Chen, Keng-Yuan
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper proposes a novel switch-mode power stage architecture for actuator array. When the number of actuator is large, the associated power amplifier becomes costly and bulky if each actuator is controlled by an independent amplifier. Further, the increased operational block raises power consumption and is inefficient. The proposed architecture allows sharing of the power ampler among actuators, and thereby reducing the number of switching devices. The switching method is controlled by an MIMO optimal feedback quantization, which is designed to minimize a weighted measure of the quantization distortion. The proposed method is applied to design a class-D stereo audio amplifier (dual actuators). The control performance and cross-talk behavior are investigated.
A novel switching strategy for FOC motor drive using multi-dimensional feedback quantization
(2012-02-01)
Chen, Keng-Yuan
;
Hu, Jwu-Sheng
;
Tang, Chi-Him
;
Shen, Te-Yang
;
電機工程學系
This work proposes a new switching strategy for field oriented control of motor driver to minimize the power of filtered error. By modifying pulse-width resolution and system operating frequency, the switching number is reduced without sacrificing the speed performance. Noise shaping induced by different sampling rates is analyzed. Conventional space vector pulse width modulation is shown to be a special case of the proposed strategy. Another case with zero-bit pulse-width resolution, MDFQM, is proposed. Experimental results show that MDFQM can reduce switching number and heat dissipation of transistors. This property enhances the control efficiency, reliability and durability of power stage. (C) 2011 Elsevier Ltd. All rights reserved.
A Procedure for Simultaneous Calibration of Environmental Cameras and Localization of Mobile Robots
(2009)
Hsieh, Hsiang-Wen
;
Wu, Chin-Chia
;
Hu, Jwu-Sheng
;
電控工程研究所
;
Institute of Electrical and Control Engineering
With an accurate estimate of the robot pose, robots will be capable of navigating in the environment autonomously and efficiently. In this paper, an infrastructure comprises of simultaneous camera calibration and robot localization is presented. The presented approach utilizes particle filter that integrates odometry data from robot and images captured from overhead cameras in the environment to localize the robot and calibrate overhead camera. An odometry-based motion model, HSV histogram-based measurement model and DLT (Direct Linear Transform) are designed to estimate robot poses and calibration cameras iteratively. Experiment results show the presented approach could calibrate the camera and estimate robot poses iteratively with reprojection errors around 210 mm in world reference plane.
A Robotic Ball Catcher with Embedded Visual Servo Processor
(2010-01-01)
Hu, Jwu-Sheng
;
Chien, Ming-Chih
;
Chang, Yung-Jung
;
Chang, Yen-Chung
;
Su, Shyh-Haur
In this work we present a robotic ball catcher with embedded visual servo processor. The embedded visual servo processor with powerful parallel computing capability is used as the computation platform to track and triangulate a flying ball's position in 3D based on stereo vision. A recursive least squares algorithm for model-based path prediction of the flying ball is used to determine the catch time and position. Experimental results for real time catching of a flying ball are presented by a 6-DOF robot arm. The percentage of success rate of the robotic ball catcher was found to be approximately 60% for the ball thrown to it from five meters away.
A robust adaptive speech enhancement system for vehicular applications
(2006-08-01)
Hu, Jwu-Sheng
;
Cheng, Chieh-Cheng
;
Liu, Wei-Han
;
Yang, Chia-Hsing
;
電控工程研究所
This work proposes and implements a novel and robust adaptive speech enhancement system, which contains both time domain and frequency domain beamformers using H-infinity filtering approach to provide a clean and undisturbed speech waveform and improve the speech recognition rate in vehicle environments. A microphone array data acquisition hardware is also designed and implemented for the proposed speech enhancement system. Mutually matched microphones are needed for traditional multidimensional noise reduction methods, but this requirement is not practical for consumer applications from the cost standpoint. To overcome this issue, the proposed system adapts the mismatch dynamics to maintain the theoretical performance allowing unmatched microphones to be used in an array. Furthermore, to achieve a high speech recognition performance, the speech recognizer is usually required to be retrained for different vehicle environments due to different noise characteristics and channel effects. The proposed system using the H-infinity filtering approach, which makes no assumptions about noise and disturbance, is robust to the modeling error in a channel recovery process. Consequently, the real vehicular experimental results show that the proposed frequency domain beamformer provides a satisfactory speech recognition performance without the need to retrain the speech recognizer.(1)
A Robust Orientation Estimation Algorithm Using MARG Sensors
(2015-03-01)
Hu, Jwu-Sheng
;
Sun, Kuan-Chun
;
電機工程學系
;
電控工程研究所
;
Department of Electrical and Computer Engineering
This paper presents a robust orientation estimation algorithm by using magnetic, angular rate, and gravity sensors. The robustness is achieved by two online methods: compensation of hard iron effect for the magnetometer and separation of the accelerometer signals into gravity projections and linear accelerations. Further, direct cosine matrix is used for the rotation matrix relative to the north, east, down frame. The fusion equations are solved using state-constrained Kalman filter. The simulation and experiment results show the effectiveness of the proposed algorithm in estimating the orientation under acceleration and hard iron influence.
A robust statistical-based speaker's location detection algorithm in a vehicular environment
(2007)
Hu, Jwu-Sheng
;
Cheng, Chieh-Cheng
;
Liu, Wei-Han
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This work presents a robust speaker's location detection algorithm using a single linear microphone array that is capable of detecting multiple speech sources under the assumption that there exist nonoverlapped speech segments among sources. Namely, the overlapped speech segments are treated as uncertainty and are not used for detection. The location detection algorithm is derived from a previous work ( 2006), where Gaussian mixture models (GMMs) are used to model location-dependent and content and speaker-independent phase difference distributions. The proposed algorithm is proven to be robust against the complex vehicular acoustics including noise, reverberation, near-filed, far-field, line-of-sight, and non-line-of-sight conditions, and microphones' mismatch. An adaptive system architecture is developed to adjust the Gaussian mixture ( GM) location model to environmental noises. To deal with unmodeled speech sources as well as overlapped speech signals, a threshold adaptation scheme is proposed in this work. Experimental results demonstrate high detection accuracy in a noisy vehicular environment.