IRAQI JOURNAL OF COMPUTERS, COMMUNICATIONS, CONTROL AND SYSTEMS ENGINEERING

The applications of an Iterative method for the restoration of color blurred image is presented. The method is an enhanced modification of the fixed-Phase Iterative algorithm (FPIA). A blurred image is enhanced by Laplace operator during the FPIA method on eachiteration. This modification is originally supported theoretically by a derivation of some iterative deblurring methods that are based on the enhanced version of the blurred image instead of the blurred image itself only. The modified fixed phase iterative algorithm (MFPIA) method is examined to restore some Gaussian- and motion-blurred color images. The restored images via this proposed method are compared with some standard iterative method called Richardson-Lucy method. From the comparison. It is appearant that the MFPIA method gives better values for signal to noise ratio (SNR) of the resuling image. The method is also better from human visual measurements point of viww with less number of iterations. In addition to that benefit the restoration by the other method result in images having shaded edges. Another advantage for the MFPIA method is that to restore a blurred image, the only thing required to be known is the blurred image itself; while a prior knowledge of the point spread function (PSF) with the original unblurred image is a must in the Richardson-Lucy methods. This makes the MFPIA method more powerful for practical applications.

This paper presents an effective method for improving the performance of speaker identification system based on the multiresolution properly of the wavelet transform, the input speech signal is decomposed into L subbands. To capture the characteristic of the vocal tract, the liner prediction code of each band (including the linear predictive code (LPC)for full band) are calculated. The feature recombination schemes combines the LPC of each band and LPC for full band in single feature vector then the Euclidean distance measure is used to perform the similarity measure between the test and reference speech. Experimental results shows that the proposed method achieve better performance than speaker identification using LPC and real cepstral coefficients.

Over time, inertial navigators drift from their preset alignments. Or, the initial alignment may have been corrupted by vehicle motion, with imperfect transfer of alignment and velocities to the navigator. Also, there may not have been enough time to perfect alignment. In such case, navigators can be benefit from aiding such as GPS.So that, the combination of GPS and INS has become increasingly common in the past few years, because the characteristics of GPS and INS are complementary.The integration between the GPS and INS leads to accurate navigation solution by overcoming each of their respective shortcomings. And to make this integration possible the difference between the GPS and INS systems in sampling rate must be solved before any integration can be work properly.

This work presents a numerical study for a new design of small size and multi-band fractal dipole antennas which is constructed by combining two fractal geometries to get the main antenna body. In this work a 3D combined fractal dipole antenna will be designed based on Hilbert and Koch curves to form the antenna geometry by using method of moment technique. Numerical simulations presented here also indicate that antenna size can be further reduced by 3D combined.

This paper presents the design and practical implementation of a multi-channel PC based logic analyzer. The analyzer has16 input channels with memory depth of 4K snapshots/channel and capture rate of up to 5 MHz. The analyzer parameters such as internal or external trigger source, falling or rising edge capture clock, state or timing measuring mode, number of pre-trigger and post-trigger snapshots, are made to be programmable and could be changed manually. The analyzer prototype consists of hardware part represented by the development of the interface ISA card and the software part that involves the kernel mode driver and the GUI program development. The developed prototype analyzer has been tested under different configuration schemes using 8085 SDK.

The images have large amount of information and they are important object in modern database system, so there is a great need for efficient image indexing key as access and retrieval tool. In database systems the index key is an attribute which has unique value for all tuples of the relation by “ name“ , “ numbers“ or any codes, this key aids to identify the tuples. In archive system a vast amount of digital images it is very necessary to find unique image index key to store it corresponding to its image and can use it to retrieval documents. So in the proposed algorithm we generate image index key which depends on extracting the key from the contents of the image itself, so such an image index key will contain the properties and characteristics of image, it is Unique, because its value depends on the value of image pixels. So it is a good method for image indexing , image verification , image ordering and checking the authenticity of retrieved image to prevent any juggle perhaps happening with the image by recomputing index keyword of retrieved image and comparing it with index keyword which is stored in database. Using shrinking method With Image Index Key Extracting algorithm (IIKE) reduce storage space of stored images.

In this paper, fuzzy brake controller is proposed. It is build by using the theory of fuzzy sets. The rules base of this controller are based on the previous knowledge and experiences in vehicle field.The working of controller reduces the vehicle velocity to achieve the safe distance in front between the vehicle and anther vehicle that traveling in front of it. So the controller prevents the slip in the tires during braking.The velocity error and distance error are used as inputs and brake force as control action output for the controller.The simulations of fuzzy controller with nonlinear brake vehicle model give acceptable responses for different cases.

Digital image watermarking has become a popular technique for authentication and copyright protection. For verifying the security and robustness of watermarking algorithms, specific attacks have to be applied to test them. This paper presents different attacks approaches which are based on the estimation of the parameters used. The exploitation of the properties of Back propagation neural network gives good results for detection.

Image identification plays a great role in industrial, remote sensing, medical and military applications. It is concerned with the generation of a signature to the image. This work proposes a dynamic program (use Neural Network) to classify the texture of human member image then identify whether the member is infected or not. The program has the ability of determining which part of that member is infected depending on the comparison between the healthy member image stored in advance with a test image. The first step is to make approximation to the image using wavelet network (Wavenet) technique. Through this technique we shall get an approximated image with reduced data. In addition, we shall get implicit information to that image. The second step is to subdivide the resultant image from the first step into 16 equally subparts then deal with each subpart as a unique image.Finally, in the third step, the minimum distance (Mahalanobias Distance) approach is employed for subpart identification. All programs are written using MATLAB VER. 6.5 package.

Direct torque control became the most popular technique for induction motor control through the last two decades, because of its simple structure, accurate and fast torque response, but it has some drawbacks such as torque and stator flux ripples. Therefore, an accurate and fast estimation of stator flux and torque values is required. In this paper a proposed model for two Multi-layer Feed-Forward Neural Network (MFFNN) to simulate and train the direct torque control data of three phase induction motor for estimation of electromagnetic torque, stator flux, and flux angle at two different sampling frequencies. The feed-forward neural networks proposed consist of three layers. The input layer consists of four neurons (stator voltages and currents) and the output layer consists of three neurons (electromagnetic torque, stator flux and flux angle). Quick back-propagation algorithm is used to train the proposed networks. Simulation model is performed using MATLAB. The results have been compared according to computation time and accuracy.