elaf jabbar abdulrazzaq altaee

Nowadays, authentication process in biometric system become most critical task with the expansive of individual information in the world. Where in many current applications, devices and commercial treatments required fingerprint identification process in order to verify the requested services. Most technologies also motivate to this direction. With the increasing of fingerprints uses, there is a need to provide a technique that able to handle the issues that exist in fingerprint acquisition and verification processes. Typically, fingerprint authenticated based on pick small amount of information from some points called Minutiae points.This approach suffers from many issues and provide poor results when the samples of fingerprints are degraded (scale, illumination,direction) changes. However, BRISK algorithm used to handle the previous issues and to extract the significant information from corner points in fingerprint. BRISK is invariant to scale, illumination, and direction changes and its able to pick large number of information when compared with minutiae points. In this paper, BRISK algorithm used based on image based approach, where current recognition matrices are developed and proposed new metrics without need for human interaction. UPEK dataset used to test the performance of proposed system, where the results show high accuracy rate in this dataset. Proposed system evaluated using FAR, FRR, EER and Accuracy and based on selected metrics the proposed system and methodology achieve high accuracy rate than others, and gives a novel modification in authentication task in biometric systems

Hiding information is a significant range of study in relevant existences including a quantity of requests. Steganography or hiding information is a discipline of implanting info into the digital cover,like text, video, voice, images etc. Several Steganography approaches have been practical in relations of forcefulness and capacity. Nevertheless, furthermost of steganography technique, which used to hide a digital message with a big size it motivation on noiselessness of cover image as well as capacity. this proposal is mention to the imperceptibly of cover image by finding the optimal block of cover image for hiding block of secrete message. Moreover, our proposal go farther to find the optimal pixel for hiding 5 bits of secrete message in the adjacent pixel of cover image. Adjacent pixel it could be for 5 bits of secret message or rotation right one bit of the 5 bits

The main purpose of steganography in current paper is to hide secret image using the Least Significant Bit (LSB) technique inside cover image so that the human eye would be unable to notice the hidden image in the cover file and it became difficult for attacker to detect it. This paper introduced comparing study between two types of images format (Gif and JPEG) used as a cover. The effectiveness of the comparison based on quality measures were done by computing RMSE, SNRrms, SNRpeak, and MAE. The results showed that the LSB steganography in JPEG images is better than GIF images.

The problem of signature recognition, which is a special case of image analysis, describes how to extracts the features efficiently and verifies the signature perfectly. Human signature can be handling as the image and can be Identified by using Elman neural network. This paper deals with the signature using global features extraction and Elman neural network in which the human signature is captured In image acquisition stage, scanner was used as image acquisition device. In image processing, many techniques are used, through which the signature is extracted from the original image, filtered from noises, and image improvements that prepare the signature to the feature extraction stage. Feature extraction stage which global features is implemented and ending with vector of values for feature extraction method that describes the signature image features. Each feature vector is fed Elman neural network which, that will characterize the signature to whom it belongs. Data base contain 500 signatures for 100 person each person have 5 signatures with different orientation, size, deviation, etc. In an Elman network, the weights from the hidden layer to the context layer are set to one and are fixed because the values of the context neurons have to be copied exactly. The initial output weights of the context neurons are equal to half the output range of the other neurons in the network. The Elman network can be trained with gradient descent back propagation and optimization methods. The accuracy of proposed method was is highly promised result and dependable.

AbstractــSince an DWT- DMWT based speech scrambler retains aconsiderable residual intelligibility in the scrambled speech; thispaper proposes a new speech-scrambling technique to remove theresidual intelligibility from the scrambled speech. The proposednew technique, based on the combination of an appropriate QPSK mapping method, can scramble speech withoutresidual intelligibility by permuting several frequencycomponents. Without residual intelligibility in the scrambledspeech, the proposed QPSK speech scrambler is secure from frequency domain attacksby eavesdroppers. In addition, the QPSK speech scrambler needs only two 2D(DWT-DMWT) operations insteadof the four required by the 1D(DWT-DMWT)based speech scrambler in systemstructure. Simulation results are also presented to show theeffectiveness of this proposed technique.

In the present paper, Gray-Scale image compression using Fast Fourier Transform ( FFT ) Algorithm and Multilayer Perceptron Network ( MLPN) bases properties are studied. To analyze the performance of the Fast Fourier Transform ( FFT ) Algorithm and Multilayer Perceptron Network ( MLPN) with the gray scale images and binary images we fixed the loss amount of the data in the compressed image ( Quality of the compressed image will be same for each function used and calculated their respective Compression Measures. ( FFT ) Algorithm with ( MLPN) gives the maximum compression for a specific type of gray scale images and better than ( FFT ) Algorithm for gray scale images. The experimental results confirm the effectiveness of the proposed ( FFT ) Algorithm with ( MLPN) .

Breast cancer is the most frequently diagnosed cancer and the leading cause of cancer death of female worldwide. Mammogram is one of the most excellent technologies currently being used for diagnosing breast cancer. Computer aided diagnosis helps the radiologists to detect abnormalities earlier than traditional procedures. In this paper, we suggested to use some of features selected to distinguish the benign and malignant breast cancer. Tumor segmented and denoising prior to classification. The accuracy of proposed system was 100%.

There are serious problem for storage sparse matrix due to west of memory used for storage the non-zero values which represent more than 90% of sparse matrix. There are many algorithms suggested for solving this problem. A new storage method for large sparse matrices was presented in this paper based on quadtree. The suggested algorithm utilized the idea of quadtree to re-represent the sparse matrix in two vectors, which reduce the required memory space for storage sparse matrix. The suggested algorithm reduced the memory space required to store the zero values to more than 85%. The algorithm compared with many algorithms and it was more efficient than almost all the previous algorithms up to our knowledge. The current algorithm characterized by less space storage need, highly speed, and easy to implement.

Speech Scrambling is defined as speech signals are altered to make them unintelligible to listeners who are not the intended recipients of the communication . The main aim of this thesis is modeling and simulating a proposed structure of Speech Scrambling based on Wavelet and Multiwavelet Transformations. The structure of speech scrambling consists of three main parts: scrambling, noisy channel and descrambling. The processes implemented in scrambling and descrambling have an inverted relation. The Speech Scrambling based on Wavelet and Multiwavelet Transformations has been examined by using Arabic, and English Analog speech signals, spoken by a man or woman . The evaluation process of the performance of the scrambled speech has been taken into consideration via the using of different noise power levels in the simulated communication channel. The scrambled speech performance has been tested by using Subjective and Objective tests. The scrambled speech has investigated one type of transformations (Wavelet and Multiwavelet ) (Daubechies 4), with one level. Using (db4) Wavelet and (db4) Multiwavelet. Each one is considered as one level, adding the noise to scrambled speech signal, with ( SNR =10 dB, SNR =20 dB, and SNR =30 dB ), for the Arabic word‘ spoken by a woman’s voice "ايلاف", the English sentence spoken by a woman’s voice " Department of Computers " , and the Arabic sentence spoken by a man’s voice "وزارة التعليم العالي والبحث العلمي " . Our results confirm that the Speech Scrambling is an effective to reconstruct the original speech signal based on Multiwavelet Transformation .