نشریه مهندسی برق و مهندسی کامپیوتر ایران
پیاپی 1 (بهار و تابستان 1382)

The segmentation of foreground regions in image sequences is the first and the most important stage in many automated visual surveillance applications; and background subtraction is a method typically used for such applications. In this method, each new frame is compared with a model of the empty scene (which we call it ‘Background’), then those regions in the image that differ significantly from the background are identified as foreground.This paper presents a new background subtraction approach. In this method, each image is divided into similar NN blocks; then, some features are extracted from every block and the history of each feature are modeled as a combination of gaussian distributions. These distributions are updated after reception of every frame information. Then the gaussian distributions of the adaptive mixture models are evaluated to determine which one most likely describes the background and each block is classified as background or foreground based on the gaussians distributions which represents its feature value most effectively.The software implementations on personal computers show accepting capability of this approach for handling intruders to the scene, objects being introduced or removed from the scene, noises and unwanted changes in the background. Also, high speed of execution and reduced memory requirements makes this approach as a suitable method for high percentage of real-time applications

This paper presents two new approaches for robust step tracking in structure uncertain nonlinear systems. The problem is first restated as a non linear optimal control infinite horizon problem, then with a suitable change of variable, the time interval is transfer to the finite horizon [0 1). This change of variable, poses a time varying problem. This problem is then transfer to measure space, and it is shown that an optimal measure must be determined which is equivalent to a linear programming problem with infinite dimension. Then, using finite horizon approximations, the optimal control law is determined as a piece wise constant function. Simulations are provided to show the effectiveness of the proposed methodology.

This paper has proposed an on-line method of partial discharges (PDs) detection. Fundamental difficulty in PD measurement is that PD signal is so minute that can be easily contaminated by huge amount of noise and this makes PD detection rather obscure. Thus, noise reduction algorithms have been extensively deployed to mitigate the noise. Among which, Digital Signal Processing (DSP) techniques are becoming more and more applicable. Compared by linear predictor and Least Mean Square (LMS), a wavelet-based noise reduction algorithm has been utilised. Some significant considerations in wavelet denoising such as selection of level of decomposition and reconstruction, mother wavelets, methods of signal extension,thresholding criterion have been discussed deeply. In order to prove the effectiveness of our algorithm, real data extracted from an 11 kV cable has been used.

One of the main parameters for overhead lines designing is conductor oscillation that has a special role for determination of conductor’s gaps. One of the kinds of oscillations is galloping which creates economic losses for utilities.In this paper, a dynamic model of single conductor overhead lines that has vertical motion is investigated and according to applicable differential equations, this model is simulated for estimation of galloping amplitude. Also by using of control theories and mathematical approximation, a new electrical model is presented. Moreover by using of nonlinear control theory, system equivalent point has been determined and the amount of damper damping coefficient for decreasing of the oscillation amplitude has been evaluated.

We present a new adaptive noise cancellation method for ECG noise cancelling in very difficult conditions. This method is based on modified variable step size LMS algorithm. We show that the proposed algorithm is more efficient and the adaption speed is higher than conventional LMS.