Search

Speech recognition is one type of technology, which make a computer to recognize the voice of words that an individual speaks through a microphone and convert it into the written text. In this paper, the proposed system for helping illiterate and blind peoples to open applications with them voices. The proposed system includes two parts; the first part is the training part while the second part is used for the testing. The system contains seven-steps; the first step is the recording of voices and the second step is voice pre-processing. The third step is the feature extraction using MFCC (Mel Frequency Cepstrum Coefficient) method that involves seven steps. The fourth step is for classification voices, there are 1400 voices samples used in training by using a Naïve Bayesian method as a classifier. The fifth step is the matching step using the Correlation Coefficient, there are 200 voices samples in testing. The sixth step was to convert voice into text and the seventh step for execution one of 20 commands. The accuracy of results from using the Naïve Bayesian algorithm in the training phase gives (100 %) while the accuracy of results in the testing phase using Correlation Coefficient gives (98%).

In this paper, fourteen similarity metrics are reviews, which will be the most important part in Diagnosis, Classification, Clustering and Recognition. Most researchers may not sure to choose which metric will be powerful and give high accuracy in them researches. Therefore, this paper will be as a guide for them to select which metric useful for them research by try one of fourteen metrics that listed in this paper and can compare one of these metrics by advantage and disadvantage of each one. In addition, there are new metrics modified to give more accuracy by testing them in some clustering application.

This paper suggest approach to solve the problem of social communication between blind and dumb by converting voices of 28 Arabic letters (ي,.........,أ) into gesture (images) by extraction features by using Mel-frequency Cepstral coefficients (MFCC)and classify the types of letters by using; J48, KNN, and Naive byes (NB). Several features are extracted from speech voice of Arabic letters voices. The dataset collected by recorded voices from twenty different persons, each person recorded ten voices for each twenty eight letters so the total dataset are 5600 voices (200 voices for each 28 letters). Mel-frequency Cepstral coefficients are extracted from 5600 voices of letters which convert the voices into a signal and extract features vector to classify later by using J48, KNN and NB algorithms, which may vary in time or speed signals. The experimental results shows that the best accuracy of speech recognition algorithm by using the J48 algorithm with a performance ratio of 100% while KNN is the 94.023% and Naive byes is the 20.012%.

In this paper, we present a new way to classify four types of images (Car accidents, Fire, Abnormal objects in street and Digs) which will be sent to four government places; Civil Defence, police station and Municipal. The classification method depends on the Content-Based Image Retrieval (CBIR), where we use a new method. In this method, we use a combination of three methods to extract features from an image; Single Value Decomposition (SVD), Edge Histogram Descriptor (EHD) and Color Auto-Correlogram for Extraction Features. You will use these features to find the closest similarities to the query image from the database images by selecting the closest 3 images, then choosing the class to which the closest two images belong to the retrieved. The combined method showed 100% accuracy in training phase and 100% test phase accuracy.

Feature extraction is the main core in diagnosis, classification, clustering, recognition, and detection. Many researchers may by interesting in choosing suitable features that used in the applications. In this paper, the most important features methods are collected, and explained each one. The features in this paper are divided into four groups; Geometric features, Statistical features, Texture features, and Color features. It explains the methodology of each method, its equations, and application. In this paper, we made acomparison among them by using two types of image, one type for face images (163 images divided into 113 for training and 50 for testing) and the other for plant images(130 images divided into 100 for training and 30 for testing) to test the features in geometric and textures. Each type of image group shows that each type of images may be used suitable features may differ from other types.

In this paper, we describe the problem of sequencing a set of n jobs on single machine was considered to minimize multiple objectives function (MOF). The objective is to find the approximate solutions for scheduling n independent jobs to minimize the objective function consists from asum of weighted number of earlyjobs and weighted number of tardy jobs with fuzzy processing time.This problem is denoted by: (1/ / 𝑁ℎ𝑗𝐸𝑗 + 𝑁𝑤𝑗 𝑇𝑗 𝑛 𝑗 =1 𝑛 𝑗 =1 ). To resolve it we proposed the Average High Ranking (AHR) method to obtain a processing time generated from fuzzy processing time, calculate the costs and reach to total penalty cost. Since our problem is Strongly NP-hard in normal form, we used Simulated Annealing. It solved the problem with up to 12000 jobs in 30 seconds.

In this paper we design an expert system to diseases diagnosis by using Fuzzy set depending on doctor’s opinions. Approach: Using fuzzy set to diseases diagnosis depending on opinion of 20 doctors, Results: it has been to diagnose three types of respiratory diseases (primary kinds of respiratory diseases) (pneumonia (PEN), tuberculosis (TB) and normal influenza (INF)), there are four symptoms X-ray, Respiratory rate (RR), Cough (CO) and Fever (F) which indicate as input of the fuzzy logic and the output will be a range of the risks and type of respiratory diseases.

In this paper we have made an algorithm that uses Wavelet-Fuzzy technique to diagnose the constitutional jaundice (Gilbert, Rotor and Dubin-Johnson syndrome). We employ the wavelet transform and fuzzy logic for analysis and classification respectively. We made experiments for 55 cases mostly for children they suffered from different kinds of the constitutional jaundice. Our algorithm is useful for disease diagnosis and can accurately to detect the kind of disease. In our algorithm we enhance the images by using appropriate enhancement methods before analyzing the images by using wavelet transform. The features obtained are passed to a fuzzy system for disease detection. Our algorithm yields more accurate results compare to the diagnosis by a doctor's eyes.

In this paper, we analysis textures of remote sensing images by taking two reference remote sensing images. We employ the wavelet transform and neural network for analysis and classification respectively. We use (symmlet5) and (cioflet1) mother functions for analyzing the two images, that contains water, forest and earth. The images are gray level and (128 times 128) size. The processing is carried out to divide each image into (16) blocks with size (32 times 32). Each block will be entered to the wavelet mother function, after trying several mother functions, we found that the (Coif1, Sym5) are the best choice. The results are passed to the features extraction (mean, standard deviation, and variance) and the output is then fed as input to the neural network(NN). Finally the result from NN with (Levenberg Marquardt (LM) algorithm) gives the type of texture (forest , earth, and water).