Abdul Majid Darsono

Nonlinear H∞ feedback control with integrator for polynomial discrete-time systems☆

Abstract This paper investigates the problem of designing a nonlinear H ∞ feedback controller for polynomial discrete-time systems with and without polytopic uncertainties. The objective is to design a controller such that the ratio between the energy of the regulated outputs and the energy of the exogenous disturbance/inputs is minimized or guaranteed to be less or equal to a prescribed value. It is well known that the state dependant or parameter dependant Lyapunov function is always chosen for synthesizing polynomial discrete-time systems. This leads the solution to be nonconvex because the Lyapunov function and the controller matrix are coupled and therefore cannot be solved by semidefinite programming (SDP). Hence, in this paper, an integrator is proposed to be incorporated into the controller structure. In doing so, the coupling of Lyapunov function and controller matrix can be eliminated effectively. This somehow simplifies the numerical solution of the problem. Then, by using SOS decomposition approach, sufficient conditions for the existence of the proposed controller are provided in terms of solvability of the state-dependent linear matrix inequalities (SDLMIs) which can be solved by SDP. A tunnel diode circuit is used to demonstrate the effectiveness of this integrator approach.

Eye closure and open detection using Adaptive Thresholding Histogram Enhancement (ATHE) technique and connected components utilisation

Eye closure detection is an important operation prior to carry out the main algorithm such as iris recognition algorithms, and eye tracking algorithms. This paper introduces a method to detect eye closure using Adaptive Thresholding Histogram Enhancement (ATHE) technique and connected component utilisation. The ATHE technique is a combination of histogram enhancement and estimation threshold technique. Firstly, in this proposed method the eye region is required to be localised. The ATHE technique enhances the eye region image then and yield the threshold value to segment the iris region. Based on the segmentation result, the connected components of binary image are used to classify the state of eye whether open or close. This classification is based on the shape and size of segmented region. The performance of the proposed technique is tested and validated by using UBIRIS, MMU and CASIA iris image database.

Performance analysis of Orthogonal - Diamond Search algorithm for Motion Estimation

In recent years, many fast Block Matching Algorithm or known as BMA have been proposed and developed in order to reduce the computational cost while maintaining the quality of the video signal. In this paper, the algorithm proposed is called Orthogonal-Diamond Search (ODS) which employs an orthogonal-shaped search pattern in the first step and then is switched into diamond-shaped search pattern for the next step. The proposed algorithm and several established algorithms namely Three Step Search (TSS), New-Three-Step Search (NTSS), Four-Step Search (4SS), Diamond Search (DS) and Hexagon-Diamond Search (HDS) are implemented using MATLAB and their performance are being compared and analyzed in terms of number of search point, peak signal-to-noise ratio (PSNR) and computational complexity. Simulation result show that the proposed algorithm can find motion vector with fever number of search points while maintains close performance of video quality with other selected algorithms.

Fusion of Low Frequency Coefficients of DCT Transform Image for Face and Palmprint Multimodal Biometrics

In this paper, we propose multimodal biometric feature fusion using alternating concatenation of DCT coefficients exist in face and plamprint images. Discrete cosine transform (DCT) is used to extract low frequency features which has high discrimination feature at the top left corner of the DCT transform image. The fuse feature vector is projected to the most principal component of eigenvector to produces low dimensional fused feature vector which contains important information about the face and palmprint images. Distance classifier is then implemented as a classifier to compute the nearest distance of test feature data point with a template to evaluate the recognition process. PolyU and FERET dataset is used to validate the propose method and the result shows fusion by using alternating concatenation of face and palmprint is able to produce a better recognition rates compare to concatenation method. The best recognition rate is 95%.