Showing papers by "Shun-Feng Su published in 2015"

Vision-Based Hand Gesture Recognition System for a Dynamic and Complicated Environment

Abstract: A vision-based hand gesture recognition system is considered in this paper. Unlike other hand gesture recognition studies, our study considers a complicated background and possible dynamic motion of hands. Thus, instead of using simple background subtraction, in this study, many problems are considered, such as detection of skin color image, detection of when a hand reaches in the field of the camera view and detection of a full palm. The proposed system consists of four stages, detection of the appearance of hands, segmentation of hand regions, detection of full palm, and hand gesture recognition. Detection of the appearance of hands is to find out when a hand appears in the front of the camera. Moreover, some morphological techniques, along with two stage skin color detection, are employed to alleviate the effect of noise. The proposed two-stage skin color detection approach is adopted from the idea of handling outliers to extract the palm from a complicated background. Following that, detection of full palm is conducted to know whether the hand reaches beyond the field of the camera view. The concept of ergonomics is employed to determine whether the hand is beyond the field of the camera view. Finally, experimental results show that the proposed system is quite promising.

Direct Adaptive Fuzzy Sliding Mode Control for Under-actuated Uncertain Systems

Abstract: The development of the control algorithms for under-actuated systems is important. Decoupled sliding mode control has been successfully employed to control under-actuated systems in a decoupling manner with the use of sliding mode control. However, in such a control scheme, the system functions must be known. If there are uncertainties in those functions, the control performance may not be satisfactory.In this paper, the direct adaptive fuzzy sliding mode control is employed to control a class of under-actuated uncertain systems which can be regarded as a combination of several subsystems with one same control input. By using the hierarchical sliding control approach, a sliding control law is derived so as to make every subsystem stabilized at the same time. But, since the system considered is assumed to be uncertain, the sliding control law cannot be readily facilitated. Therefore, in the study, based on Lyapunov stable theory a fuzzy compensator is proposed to approximate the uncertain part of the sliding control law. From those simulations, it can be concluded that the proposed compensator can indeed cope with system uncertainties. Besides, it can be found that the proposed compensator also provide good robustness properties.

Deep learning and its application to general image classification

Abstract: Deep learning has recently exhibited good performance in many applications. The convolution neural network is an often-used architecture for deep learning and has been widely used in computer vision and audio recognition, and outperformed other related handcraft designed feature in recent years. These techniques compared to other artificial intelligence algorithms and handcraft features need extremely much more time in training and testing and then were not widely used in the early days. Our study is about the impacts of different factors used in the convolution neural network. The considered factors are network depth, numbers of filters, and filter sizes. The used data set is the CIFAR dataset. According to our experiments, some suggestions about those factors are recommended in this study.

Wall following and continuously stair climbing systems for a tracked robot

Abstract: Autonomous mobile robots have been developed in many types and functions so far. This paper mainly presents the functionalities of self-made tracked robots and its practicality. The robot can move, patrol, and cross floors autonomously in indoor environments. With the information gotten from the Xtion sensor, the movement modes developed in this paper help the tracked robot identify different environments and successfully complete patrolling and climbing.

Design of fuzzy-based magnetic suspension vibrator for electric wheelchair

Abstract: A fuzzy-based magnetic suspension vibrator (FMSV) is designed and embedded into electrical wheelchair to promote its shock absorbing ability. Magnetic levitation system (MLS), using the electromagnetic force to float, rotate or move a suspended object, is a non-contact operation mode. This mode can effectively reduce the mechanical vibrations, friction and wearing loss caused by contact operation. Furthermore, it can also prolong the equipment's life, reduce maintenance frequency and noise. However, the MLSs possess the characteristics of nonlinear and dynamic properties, so a robust adaptive fuzzy controller (AFC) for magnetic suspension vibrator is proposed in this study. With the help of magnetic suspension vibrator, the electric power wheelchair can steer in a bumpy road and provide more comfortable riding environment for disabled people or patients.

Adaptive PD fuzzy control with dynamic learning rate for two-wheeled balancing six degrees of freedom robotic arm

Abstract: The paper proposes a novel method for enhanced fusion adaptive fuzzy control in the two-wheeled balancing six degrees of freedom robotic arm The two-wheeled mobile robot system and six degrees of freedom robotic arm are integrated into the mobile robot system Due to the motion of the robot arm, the stability issue becomes more complex This study employs the adaptive fuzzy control to provide suitable controller for the mobile robot system In addition, a dynamic learning rate can effectively improve the learning performance In order to show the superiority of the proposed controller, this paper compares our proposed method with other control methods as proportional-derivative (PD) controller, fuzzy controller, adaptive fuzzy controller, PD fuzzy controller, and adaptive PD controller The experiment results clearly demonstrate that the proposed control method has much faster convergent speed and is very well performed

Adaptive intelligent steering control of ball-riding human transporter

Abstract: This paper presents an adaptive intelligent steering controller using backstepping sliding-mode control and recurrent wavelet fuzzy cerebellar model articulation controller (RWFCMAC) and for an uncertain ball-riding human transporter in presence of significant system uncertainties. The proposed controller operates at two independent modes: self-balancing and station keeping. The self-balancing mode is used to balance by following the rider's two-dimensional inclination angles, while the station-keeping mode is aimed to permit the rider to keep the vehicle at a target position. The RWFCMAC is designed to online learn the uncertainties caused by riders' weights and different parameters. The superior performance and merit of the proposed control methods are well exemplified by conducting two simulations.