Journal of ICT Research and Applications
Not a member yet
359 research outputs found
Sort by
Free Model of Sentence Classifier for Automatic Extraction of Topic Sentences
Get PDFThis research employs free model that uses only sentential features without paragraph context to extract topic sentences of a paragraph. For finding optimal combination of features, corpus-based classification is used for constructing a sentence classifier as the model. The sentence classifier is trained by using Support Vector Machine (SVM). The experiment shows that position and meta-discourse features are more important than syntactic features to extract topic sentence, and the best performer (80.68%) is SVM classifier with all features.
Edge Connectivity Problems in Telecommunication Networks
Get PDFIf a communication network N with n stations has every station linked with at least [n/2] other stations, then the edge-connectivity of N equals its minimum degree. Also, in general, this limitation is stated to be the best possibility, as was proved by Chartrand in 1966. A more developed notion of edge-connectivity is introduced, which is called k-component order edge connectivity. It is the minimum number of edges required to be removed so that the order of each disconnected component is less than k
A Proposed Arabic Handwritten Text Normalization Method
No full textText normalization is an important technique in document image analysis and recognition. It consists of many preprocessing stages, which include slope correction, text padding, skew correction, and straight the writing line. In this side, text normalization has an important role in many procedures such as text segmentation, feature extraction and characters recognition. In the present article, a new method for text baseline detection, straightening, and slant correction for Arabic handwritten texts is proposed. The method comprises a set of sequential steps: first components segmentation is done followed by components text thinning; then, the direction features of the skeletons are extracted, and the candidate baseline regions are determined. After that, selection of the correct baseline region is done, and finally, the baselines of all components are aligned with the writing line. The experiments are conducted on IFN/ENIT benchmark Arabic dataset. The results show that the proposed method has a promising and encouraging performance
A Fast and Efficient Thinning Algorithm for Binary Images
No full textSkeletonization "also known as thinning" is an important step in the pre-processing phase in many of pattern recognition techniques. The output of Skeletonization process is the skeleton of the pattern in the images. Skeletonization is a crucial process for many applications such as OCR and writer identification. However, the improvements in this area are only a recent phenomenon and still require more researches. In this paper, a new skeletonization algorithm is proposed. This algorithm combines between parallel and sequential, which is categorized under an iterative approach. The suggested method is conducted by experiments of benchmark dataset for evaluation. The outcome is to obtain much better results compared to other thinning methods that are discussed in comparison part
Modeling Marine Electromagnetic Survey with Radial Basis Function Networks
Get PDFA marine electromagnetic survey is an engineering endeavour to discover the location and dimension of a hydrocarbon layer under an ocean floor. In this kind of survey, an array of electric and magnetic receivers are located on the sea floor and record the scattered, refracted and reflected electromagnetic wave, which has been transmitted by an electric dipole antenna towed by a vessel. The data recorded in receivers must be processed and further analysed to estimate the hydrocarbon location and dimension. To conduct those analyses successfuly, a radial basis function (RBF) network could be employed to become a forward model of the input-output relationship of the data from a marine electromagnetic survey. This type of neural networks is working based on distances between its inputs and predetermined centres of some basis functions. A previous research had been conducted to model the same marine electromagnetic survey using another type of neural networks, which is a multi layer perceptron (MLP) network. By comparing their validation and training performances (mean-squared errors and correlation coefficients), it is concluded that, in this case, the MLP network is comparatively better than the RBF network[1].[1] This manuscript is an extended version of our previous paper, entitled Radial Basis Function Networks for Modeling Marine Electromagnetic Survey, which had been presented on 2011 International Conference on Electrical Engineering and Informatics, 17-19 July 2011, Bandung, Indonesia
Robust Reed Solomon Coded MPSK Modulation
Get PDFIn this paper, construction of partitioned Reed Solomon coded modulation (RSCM), which is robust for the additive white Gaussian noise channel and a Rayleigh fading channel, is investigated. By matching configuration of component codes with the channel characteristics, it is shown that this system is robust for the Gaussian and a Rayleigh fading channel. This approach is compared with non-partitioned RSCM, a Reed Solomon code combined with an MPSK signal set using Gray mapping; and block coded MPSK modulation using binary codes, Reed Muller codes. All codes use hard decision decoding algorithm. Simulation results for these schemes show that RSCM based on set partitioning performs better than those that are not based on set partitioning and Reed Muller Coded Modulation across a wide range of conditions. The novel idea here is that in the receiver, we use a rotated 2^(m+1)-PSK detector if the transmitter uses a 2^m-PSK modulator
Determining the Standard Value of Acquisition Distortion of Fingerprint Images Based on Image Quality
Get PDFThis paper describes a novel procedure for determining the standard value of acquisition distortion of fingerprint images. Knowledge about the standard value of acquisition distortion of the fingerprint images is very important in determining the method for improving image quality. In this paper, we propose a model to determine the standard value that can be used in classifying the type of distortion of the fingerprint images based on the image quality. The results show that the standard value of acquisition distortion of the fingerprint images based on the image quality have values of the local clarity scores (LCS) follows: dry parameter values are in the range of 0.0127-0.0149, neutral parameter values are less than 0.0127, and oily parameter values are greater than 0.0149. Meanwhile, the global clarity scores (GCS) are as follows: dry parameter values are in the range of 0.0117-0.0120, neutral parameter values are less than 0.0117, and oily parameter values are greater than 0.0120; and ridge-valley thickness ratios (RVTR) are as follows: dry parameter values are less than 7.75E-05, neutral parameter values are 7.75E-05-5.94E-05, and oily parameter values are greater than 5.94E-05
Ultrafuzziness Optimization Based on Type II Fuzzy Sets for Image Thresholding
Get PDFImage thresholding is one of the processing techniques to provide high quality preprocessed image. Image vagueness and bad illumination are common obstacles yielding in a poor image thresholding output. By assuming image as fuzzy sets, several different fuzzy thresholding techniques have been proposed to remove these obstacles during threshold selection. In this paper, we proposed an algorithm for thresholding image using ultrafuzziness optimization to decrease uncertainty in fuzzy system by common fuzzy sets like type II fuzzy sets. Optimization was conducted by involving ultrafuzziness measurement for background and object fuzzy sets separately. Experimental results demonstrated that the proposed image thresholding method had good performances for images with high vagueness, low level contrast, and grayscale ambiguity
A 2.3/3.3 GHz Dual Band Antenna Design for WiMax Applications
Get PDFA triangular-slot antenna with rectangular patch for 2.3/3.3 GHz WiMax applications has been implemented on DICLAD 527 substrate (relative permittivity εr = 2.5) with 1.524 mm of substrate thickness. A rectangular patch printed on one side of the substrate is fed by a 50 Ω microstrip line and acts as the frequency tuning stub, while the triangular slot is positioned on the opposite side of the substrate, center lined to the rectangular stub. From the measurement results based on VSWR = 2 or equal to the return loss of 9.53 dB, at the lower band of 2.3 GHz the resulting impedance bandwidth is 290 MHz (from 2.16 to 2.45 GHZ) and at the upper band of 3.3 GHz is 370 MHz (from 3.31 to 3.68GHz), providing services for 2.3 GHz and 3.3 GHz frequency bands allocated for WiMax applications. The antenna gain measurement at 2.3 GHz frequency band is almost agrees with the simulation result of 3.2 dBi. While at 3.3 GHz the gain is approximately 4.4 dBi and continues to decrease with increasing frequency. The antenna gain measurement achieves maximum of 4.8 dBi (6 dBi from simulation) at about 3 GHz. The simulation and measurement results are evaluated and discussed