Mahmood Al-khassaweneh

Fault Diagnosis in Internal Combustion Engines Using Extension Neural Network

The internal combustion engine (ICE) is a special type of reciprocating and rotating machine which is an essential part of every automobile and industry in our modern life. Various faults frequently encounter this machine and cause significant losses. Thus, in this paper, we propose an effective and automated technique to diagnose the faults. Unlike the existing methods in this field, the emitted sound signal of the “ICE” is exploited as the information carrier of the faults, wavelet packet decomposition is used as the feature extraction tool, and finally, extension artificial neural network is used for the classifications of the extracted features. The extension neural network (ENN) consists of just the input layer and the output layer. This simple structure of the “ENN” enhances the performance compared to the traditional neural networks and enables us to easily insert any new information, like a new fault or new feature. Therefore, “ENN” is adaptive for new information by just adding new nodes without affecting the previously built network. The results of the proposed method show the effectiveness and the high recognition rate in classifying different faults.

A time-frequency inspired robust image watermarking

With the development of the Internet and the multimedia tools, reproduction and redistribution of digital data, such as images, audio and video have become easier than ever. This ease of access to digital data brings with itself the problem of copyright protection. Watermarking has been proposed as a solution to this problem. Two major methods have been used for watermarking; spatial and spectral domain methods. In this paper, we introduce a new framework for watermarking digital images that is inspired by the joint spatial-spectral transforms such as the Wigner distribution. The embedding and the detection algorithms in this joint domain are defined. The robustness of the proposed algorithm under attacks is illustrated.

Spatially Adaptive Wavelet Thresholding for Image Watermarking

In this paper, we introduce a new robust image watermarking technique based on the discrete wavelet transform (DWT). The proposed method extends the concept of image denoising to watermarking. A spatially adaptive wavelet thresholding method is used to select the coefficients to be watermarked. A multi-bit watermark is embedded into the discrete wavelet coefficients of the host image. A semi-blind watermark extraction algorithm is presented and the threshold for a given probability of false alarm is derived. The simulation results show that the proposed method outperforms a well-known DWT based watermarking method under most attacks including JPEG compression

The Relationship Between Two Directed Information Measures

Two directed information measures have been used extensively in literature to evaluate the direction of information flow between two time series. The relationship between these two measures, however, has not been established as to date. In this letter, we derive a formula that defines the relationship between these two measures. We also offer some insights into the interpretation of these measures and their relationship, which are verified through simulations.

Fault diagnosis in vehicle engines using sound recognition techniques

Vehicle engine faults are serious faults that occur inside the engine, the ability to successfully perform fault diagnosis is highly dependent on technician skills. Some experienced technicians have some failure rate, which can lead to serious waste in time and money. Accordingly, an improved diagnosing methods is highly needed. In this paper, we develop an algorithm for fault diagnosis in vehicle engines using sounds techniques, since each engine fault has a specific sound that is distinguished. We collect and analyze sound samples from different types of cars, which represent different types of fault, to create a database of sound prints that will make the whole process of diagnosing engine faults based on sound easier and less time consuming. Different features from the sound samples are extracted and used for diagnosis. The fault under test is compared with the faults in the database according to their correlation, normalized mean square error, and formant frequencies values. The best match is considered the detected fault. The developed system can be useful for the inexperienced technicians and engineers and can be used as a training module for them. The simulation results show the high fault detection rates of the proposed algorithm.

Human face detection using skin color information

This paper implements a new method to detect a human skin and faces from colored images. The proposed system based on the detection of all pixels in colored images which are probably a human skin via a reference skin colors matrix. The image then goes through some modifications to enhance the face detection. The circularity feature was used to distinguish human faces from other objects with similar skin color. The proposed system was tested using MatLab using different real images and the simulation results show effectiveness of the proposed method.

Image encryption scheme based on using least square approximation techniques

One of the main application of image encryption is to transmit secure information between parties. The aim is to transmit the image securely over the network such that no unauthorized user be able to decrypt the image. In this paper, we propose a new encryption algorithm by transforming the original image into the encrypted one using randomly generated vectors. The original image is decrypted by applying least square approximation techniques on the encrypted image and the randomly generated vectors. The method has been tested for a large number of images. The numerical results have demonstrated the effectiveness of the proposed algorithm and shown enhancement in security.

A high dimensional Directed information estimation using data-dependent partitioning

Directed Information (DI) is used to quantify the causal and dynamic relations between two signals. The main advantage of using DI compared to other measures of causality is that it does not assume an underlying signal model and thus can capture both linear and nonlinear interactions between signals. However, one major problem in computing the DI from data is the high computational cost and the unreliability of the probability density function (pdf) estimation methods. In this paper, we propose a high dimensional DI estimation method based on computing multi-information by an adaptive datadependent partitioning technique. The proposed estimation method does not assume any distribution for the data under consideration and requires no pdf estimation. The proposed method is applied on simulated data and is compared with other DI estimation methods to verify its effectiveness.

Asthma attack monitoring and diagnosis: A proposed system

In recent decades, childhood asthma has become more widespread and thus become a worldwide concern. The diagnosis of childhood asthma depends highly on the physician experience and the collaboration from the child part, which sometimes hard to achieve. In many cases, the symptoms might mislead in the diagnosis of asthma. In this paper, we propose a method to diagnose and classify asthma in children. The proposed method is based on using cough sound to extract features that help in asthma diagnosis. Moreover, we propose a hardware system for asthma attack monitoring. The proposed system was implemented and performed by a self-developed computer program written in MATLAB using many cough sound samples of asthmatic and non-asthmatic children.

A value transformation and random permutation-based coloured image encryption technique

With the recent advances in multimedia transmission and storage, security becomes an important issue to protect data from unauthorised access. There are several methods proposed to secure data including watermarking, encryption, steganography and others. Image encryption, which is widely used in many applications to provide high levels of security such as internet communication, multimedia systems and medical imaging, is proposed in this paper. The proposed encryption method is based on value transformation in which the image pixels are transformed to new values and random permutations in which the rows and the columns of the image are shuffled. The value transformation step is applied on pixels values through performing logical AND operation between their values and specific value in order to get new transformed image with new pixels values unrelated to the original pixels. Moreover, the random permutation step is applied to the columns and rows of the transformed image. In this step the rows and the columns are shuffled respectively using random permutation key to produce the encrypted image. The same generated key is used to shuffle both the rows and the columns of the image. The simulation results show the effectiveness and the security of the proposed method.