Ziad Osman

Automatic processing of Arabic text

Automatic recognition of printed and handwritten documents remains an active area of research. Arabic is one of the languages that present special problems. Arabic is cursive and therefore necessitates a segmentation process to determine the boundaries of a character. Arabic characters consist of multiple disconnected parts. Dots and Diacritics are used in many Arabic characters and can appear above or below the main body of the character. In Arabic, the same letter has up to four different forms depending on where it appears in the word and depending on the letters that are adjacent to it. In this paper, a novel approach is described that recognizes Arabic script documents. The method starts by preprocessing which involves binarization, noise reduction, and thinning. The text is then segmented into separate lines. Characters are then segmented by determining bifurcation points that are near the baseline. Segmented characters are then compared to prestored templates to identify the best match. The template comparisons are based on central moments, Hu moments, and Invariant moments. The method is proven to work satisfactorily for scanned printed Arabic text. The paper concludes with a discussion of the drawbacks of the method, and a description of possible solutions.

Error Correction of Noisy Block Cipher Using Cipher and Plaintext Characteristics

Contemporary proven cryptographic algorithms, like the Advanced Encryption Standard (AES), are used in many secure data storage systems. Cipher data when written or read might be subject to noise. Classical error detection and correction methods are not suitable for encrypted data. In this paper, error detection and correction is performed at the receiver end, without any changes to the encryption algorithm. One of the properties of encrypted information is that all encrypted blocks have a minimum hamming distance from each other. This property is exploited to obtain the exact correct block. When error correction based on the encrypted data cannot be performed, natural language properties of plaintext data are used to eliminate noise. The plaintext blocks surrounding the noisy plaintext block are used to generate possible candidates. In case a unique solution is not achieved, n-gram properties of the plaintext language are used to rank the possibilities and promote the best fit.

Iris Recognition Using Phase Congruency

The techniques used in biometrics are diverse depending on which organ is of interest. When dealing with iris recognition some work was previously done. This paper describes a technique for iris recognition using phase congruency. Phase congruency is used to identify the different feature types that are present in the rich iris texture. It provides an illumination and contrast invariant measure of the significance of feature points. In this paper we will demonstrate how to use phase congruency to locate and identify the minutiae of the iris. An Euclidean distance classifier was used for comparison. Results of this method are seen to be of practical use.

A New Modified Circular Disc Dipole Antenna for Ultra-Wideband Systems

In this paper, we present a modified ultra-wideband (UWB) antenna for ultra-wideband applications such as Wireless Communication and WiMax systems. It consists of a two circular-disc patch fed by a 50 Omega coaxial feed line. The obtained bandwidth (SWR les 2) is about 10 GHz (from 0.650 to 10 GHz with a bandwidth ratio (1:15.4)). The radiation characteristics are as expected for dipole antennas.

Correcting Noise in Block Ciphers

Contemporary proven block cryptographic algorithms, like the Advanced Encryption Standard (AES), are used for secure transmission of data in many of the communication systems in use. Classical error detection/correction methods are not suitable for such situations due to the randomness property of encryption. A one bit error in an encrypted message will lead to an average of 50% error rate in the decrypted message. In this paper, error detection and correction is done at the receiver end, without any changes to the encryption algorithm. Errors that are not correctible by existing error correction methods are treated in this paper by looking at their surrounding data. For a specific corpus the space of all possible messages is modeled. The neighborhood of the noise contaminated data is used, along with the space of all possible messages, in order to correct the errors. In case of multiple possible corrections, statistical natural language processing methods are used to rank the possibilities and choose the best one.

Novel speed detection scheme

This paper shows a new scheme for vehicle speed detection. The LNB (Low Noise Block) used in mw satellite receivers is modified to act as the heart of the hand gun transceiver. An original control circuitry is suggested to make the radar detector user friendly and to inform the vehicle of its speed status. Finally, the system is implemented and tested with acceptable performance.

Emotion recognition in Arabic speech

In this paper, emotion recognition in Arabic spoken data is studied for the first time. A realistic speech corpus from Arabic TV shows is collected. The videos are labeled by their perceived emotions; namely happy, angry or surprised. Prosodic features are extracted and various classification methods are applied. Results are analyzed in this paper and conclusions and future recommendations are identified.

AGC with signal offset and peak-to-peak amplitude stabilization through feedback control

In modern electronic systems, signal stability is a crucial issue. Many methods have been developed to mitigate the deviation from an ideal signal. One of them is called Automatic Gain Control, which tracks variation in an input signal to settle output amplitude. This paper proposes a Differential Automatic Gain Controller with offset stability. It is a Proportional Integral Derivative controller that stabilizes the gain and offset simultaneously. The proposed model is simulated using Pspice, and the simulation shows that accurate results are obtained across frequencies from 1KHZ to 10GHZ. It also indicates that the error tolerance is 0.08% for amplitude and 0.38% for offset in the worst case.