Search

To reduce the manpower and response time for surveillance systems at low cost, in this paper, an ARM-based embedded system dedicated for unattended realtime moving target detection is constructed. The comprehensive procedures in building up an embedded system such as setup environment for cross-compilation, migration of Bootloader, migration of Linux-2.6 kernel, fabrication and migration of root document system and setup of peripheral driving devices have been presented. The algorithm of image background subtraction for moving target detection and tracking technology has been presented. In this embedded system, for two consecutive 640×480 image frames captured by camera, if the difference of 8-bit gray level value of a pixel at the same position is greater than 32, that pixel is marked as a “moving pixel”. When there are more than 500 moving pixels in two consecutive image frames, the camera is triggered to take pictures because the system assumes that a moving “invader” appears. Consequently, the system will transfer the taken invader's pictures to “The Cloud” through WiFi to prevent the pictures being destroyed by the invader. The constructed embedded system can be used in a security system and other applications with proper modifications.

The purpose of this paper is to design an intelligent wheel robot, which can recognize and follow a predefined forward sign while automatically bypassing any encountered obstacle. By distributing those forward signs, the path of the robot is determined. With this concept, an image based autopilot system with immunity against electromagnetic interference is constructed. Harr-like feature training algorithm in OpenCV has been applied so that the robot can quick detect the forward sign. The rotation of the robot for automatic target detecting is achieved by fuzzy control. The experimental results showed that the robot could successfully detect forward sign and response properly. Consequently, the proposed design approach is proven to be workable. Simply redistributing the recognizable signs by the robot, a new path for robot is constructed. Therefore, it has great flexibility for applications. The detail will be presented in part II.

The purpose of this paper is to design an intelligent wheel robot, which can recognize and follow a predefined forward sign while automatically bypassing any encountered obstacle. By distributing those forward signs, the path of the robot is determined. With this concept, an image based autopilot system with immunity against electromagnetic interference is constructed. Harr-like feature training algorithm in OpenCV has been applied so that the robot can quick detect the forward sign. The rotation of the robot for automatic target detecting is achieved by fuzzy control. The experimental results showed that the robot could successfully detect forward sign and response properly. Consequently, the proposed design approach is proven to be workable. Simply redistributing the recognizable signs by the robot, a new path for robot is constructed. Therefore, it has great flexibility for applications.

This paper addresses the need of the novel reciprocal state space (RSS) framework to be the supplement of standard state space system in control designs. A model following design based on variable structure control design approach utilizing direct state derivative measurement for an RSS system with nonlinear dead-zone input is proposed. Simulations for both nonlinear dead-zone input and nonlinear sector input are carried out for an unstable RSS system to successfully verify the proposed algorithm. It shows that the design procedure is straightforward for systems with direct state derivative measurement if we express the system in RSS framework.

There are two major problems that people usually have to face when they carry out optimal control designs in state space system or descriptor system. One is no state derivative considered in cost functional, the other is that some systems cannot easily design optimal control in either state space system or generalized system frameworks. To overcome the problems, a novel state derivative space system (SDS system) framework is proposed. In this novel framework, novel differential Lagrange multiplier should be used to adjoin SDS system constraint to the cost functional that is function of state derivative to straightforwardly carry out optimal control design. It showed that some singular systems with impulse mode can easily carry out optimal control in SDS system. The purpose of this research is to develop novel and simple optimal control algorithms in SDS system so that wider ranges of problems can be solved without much of mathematics overhead.

This paper proposes an anisotropic probabilistic neural network image interpolation (APNNI2). The method used anisotropic Gaussian kernel to improve probabilistic neural network image interpolation (PNNI2) which caused blurring edge. The method is better for sharpness enhancement at edge region. We interpolate slanted-edge image to observe blurring and blocking effect. Finally, we report the performance of these newly proposed methods in other image interpolation method.

This paper investigates the novel variable structure control (VSC) designs with state derivative feedback in nontraditional reciprocal state space (RSS) framework. Novel switching functions based on the derivative of sliding surface satisfying the approaching condition of sliding mode are introduced. In addition, algorithm of finding upper bound of system uncertainty is developed for robustness analysis. VSC design is carried out for unstable systems with system uncertainty and matched external disturbance. Simulation results successfully verify the validities of the proposed algorithms.

This paper introduces novel tracking control algorithms for step command utilizing state derivative output feedback and feedforward in reciprocal state space framework. Linear quadratic regulator (LQR) design of deviation system is used to find state derivative output feedback gain and two methods are developed to find feedforward gain to construct tracking controller in reciprocal state space framework. For the systems with state derivative sensors, applying the proposed algorithms, tracking control designs become straightforward while unnecessary integrations can be avoid to reduce implementation cost. Simulations of a nano-positioner with micro-accelerometer as sensor have been successfully carried out to verify the proposed algorithms. From the simulation results, we find that although the design is based on step response shaping, it also gives good time responses for time varying reference commands such as sine wave and square wave.

The paper describes the criteria of stability, controllability and observability for the systems, which can be expressed in the 'reciprocal state space framework'. Several important reasons are presented to justify the need for the 'reciprocal state space framework' and advantages of the 'reciprocal state space framework' are explained and contrasted with those of the typical standard state space framework. This concept of the direct state derivative feedback control design in the 'reciprocal state space framework' is new and attractive, especially for descriptor systems, singular systems or semi-state systems.

Introduces a design approach based on a new framework called 'reciprocal state space' framework in which the state derivative can be sensed and fed back directly. By using accelerometers and integrators, information of acceleration, velocity and displacement can be obtained and fedback. The proposed methodology using the reciprocal state space framework is useful in structural control because accelerometers are widely used in structural control systems. The proposed methodology is is illustrated by considering the vibration control problem for an aircraft wing, modeled by the finite element method.

In this paper, a design approach for controlling a composite plate structure subject to aeroelastic loading using piezoelectric actuators and sensors is presented. The nature of the sensing variables is exploited and accommodated in the control design algorithm. The controller is a state estimate based feedback using various measurements related to motion variables such as accelerations, velocities and displacements. Controller designs are based on equivalent first-order state-space baseline design gains that use existing algorithms. The proposed method provides different norms for estimator and controller gains, thereby allowing more flexibility in gain magnitudes and selection of sensors.