Showing papers in "WSEAS Transactions on Signal Processing archive in 2009"

Fast copy-move forgery detection

Abstract: This paper proposes a method for detecting copy-move forgery over images tampered by copy-move. To detect such forgeries, the given image is divided into overlapping blocks of equal size, feature for each block is then extracted and represented as a vector, all the extracted feature vectors are then sorted using the radix sort. The difference (shift vector) of the positions of every pair of adjacent feature vectors in the sorting list is computed. The accumulated number of each of the shift vectors is evaluated. A large accumulated number is considered as possible presence of a duplicated region, and thus all the feature vectors corresponding to the shift vectors with large accumulated numbers are detected, whose corresponding blocks are then marked to form a tentative detected result. Finally, the medium filtering and connected component analysis are performed on the tentative detected result to obtain the final result. Compared with other methods, employing the radix sort makes the detection much more efficient without degradation of detection quality.

A cumulant-based method for gait identification using accelerometer data with principal component analysis and support vector machine

Abstract: In this paper a cumulant-based method for identification of gait using accelerometer data is presented. Acceleration data of three different walking speeds (slow, normal and fast) for each subject was acquired by the accelerometer embedded in cell phone which was attached to the person's hip. Data analysis was based on gait cycles that were detected first. Cumulants of order from 1 to 4 with different number of lags were calculated. Feature vectors for classification were built using dimension reduction on calculated cumulants by principal component analysis (PCA). The classification was accomplished by support vector machines (SVM) with radial basis kernel. According to portion of variance covered in the calculated principal components, different lengths of feature vectors were tested. Six healthy young subjects participated in the experiment. The average person recognition rate based on gait classification was 90.3±3.2%. A similarity measure for discerning different walking types of the same subject was also introduced using dimension reduction on accelerometer data by PCA.

Perceptual distortion metric for stereo video quality evaluation

Abstract: Stereo video is regarded as an important developing trend of video technology and there is an increasing need to develop efficient and perceptually consistent methods for stereo video quality evaluation in the fields of stereo video signal processing. In this paper, a perceptual metric for stereo video quality evaluation is proposed based on the state-of-the-art physiological and psychological achievements on human visual system (HVS). Several main HVS properties related to stereo video are analyzed and a multi-channel vision model based on 3D wavelet decomposition is proposed. Simulations are performed and experimental results reveal that, compared with the traditional objective metrics such as peak signal-to-noise ratio (PSNR) and mean squared error (MSE), the proposed metric is more perceptually consistent.

Multi-class support vector machine classifier in EMG diagnosis

Abstract: The shapes of motor unit action potentials (MUAPs) in an electromyographic (EMG) signal provide an important source of information for the diagnosis of neuromuscular disorders. In order to extract this information from the EMG signals recorded at low to moderate force levels, it is required to: i) identify the MUAPs composed by the EMG signal, ii) cluster the MUAPs with similar shapes, iii) extract the features of the MUAP clusters and iv) classify the MUAPs according to pathology. In this work, three techniques for segmentation of EMG signal are presented: i) segmentation by identifying the peaks of the MUAPs, ii) by finding the beginning extraction point (BEP) and ending extraction point (EEP) of MUAPs and iii) by using discrete wavelet transform (DWT). For the clustering of MUAPs, statistical pattern recognition technique based on euclidian distance is used. The autoregressive (AR) features of the clusters are computed and are given to a multi-class support vector machine (SVM) classifier for their classification. A total of 12 EMG signals obtained from 3 normal (NOR), 5 myopathic (MYO) and 4 motor neuron diseased (MND) subjects were analyzed. The success rate for the segmentation technique used peaks to extract MUAPs was highest (95.90%) and for the statistical pattern recognition technique was 93.13%. The classification accuracy of multi-class SVM with AR features was 100%.

Image restoration viawiener filtering in the frequency domain

Abstract: In this paper, first, the performance of the Wiener filter in the frequency domain for image restoration is compared with that in the space domain on images degraded by white noise After finding that the Wiener filter in the frequency domain is more effective than that in the space domain in an ideal case, power spectrum estimation methods for the Wiener filter in the frequency domain are discussed Three approaches are considered; frequency band division processing (FBDP), modified FBDP and averaging high frequency components (AHFC) The performances of the Wiener filter with the three approaches for power spectrum estimation are investigated through computer simulation experiments It is shown that the frequency domain Wiener with the modified FBDP provides a superior performance relative to that with the FBDP and AHFC

Experiments in room acoustics: modelling of a church sound field and reverberation time measurements

Abstract: In this paper a study of the acoustical response of a new built church is shown. This study is based on the measurement of reverberation time, adopting the noise interrupted method, according to the International Standard. This method allows to evaluate the reverberation time by means of acoustical sound level acquisition and analysis. The reverberation time is one of the principal parameters to be optimized in order to design and/or verify the acoustical behaviour of a room and consequently to guarantee a good people hearing sensation. In a post-opera intervention, the reverberation time can be improved modifying the reflecting surfaces of walls, floor and roof, in order to reduce the energetic contributions of late reflections. This improvement can be achieved by replacing or covering reflecting surfaces with absorbing panels or carpets. The design of an appropriate intervention can be aided by a dedicated simulation software, as it is shown in the last part of the paper.

Noisy image restoration based on boundary resetting BDND and median filtering with smallest window

Abstract: In this paper, a restoration approach for noisy image is proposed where a boundary resetting boundary discriminative noise detection (BRBDND) and a median filtering with smallest window (MFSW) are applied. In the proposed image restoration approach, two stages are involved: noise detection and noise replacement. The BRBDND is used to detect noisy pixels in an image. If a pixel is uncorrupted, then keep it intact. Or replace it with an uncorrupted neighborhood pixel through the MFSW. Note that miss detection happens in the BDND presented in [17] when the noise density is high. The miss detection is even worse for cases with unbalanced noisy density where the portions for the salt noise and the pepper noise are different. A boundary resetting scheme is incorporated into the BDND. By this doing, the problem of miss detection described above can be prevented. Note that a larger window used in the median filtering leads to a stronger smoothing effect on the restored image. The reported median filtering approaches, like the modified noise adaptive soft-switching median filter (MNASM) in [17], uses larger windows generally. Thus, a median filtering with smallest window (MFSW) is proposed to improve the visual quality of restored image. Two examples are provided to justify the proposed image restoration approach BRBDND/MFSW where comparisons are made with the BDND/MNASM. The results indicate that the proposed BRBDND is able to deal with the miss detection problem in the BDND. It also shows that the proposed MFSW indeed improves the visual quality of restored image as expected. The simulation results suggest that the proposed restoration approach BRBDND/MFSW generally outperforms the BDND/MNASM both in the PSNR and the visual quality of restored image.

Wavelet filter design based on the lifting scheme and its application in lossless image compression

Abstract: The description of filter banks using lifting structures does not only benefit low-complexity implementation in software or hardware, but is also advantageous for the design of filter banks because of the guaranteed perfect reconstruction property This paper proposes a new design method for wavelet filter banks, which is explained based on a single lifting structure suitable for 9/7 filter pairs The filters are derived directly, the factorisation of known filters is not necessary In addition, it is shown that the signal boundaries can be treated with little computational efforts The modification of the standard design constraints leads to families of related filter pairs with varying characteristics It includes a filter bank that can be implemented in integer arithmetic without divisions, shows better performance than the standard 9/7 filter bank for lossless image compression and competitive performance when applied in lossy compression

Fast image matching on web pages

Abstract: In this paper, a fast method for image matching on web pages is presented. Such method relies on performing cross correlation in the frequency domain between the web image and the image given in the user query. The cross correlation operation is modified. Instead of performing dot multiplication in the frequency domain, image subtraction is applied in two dimensions. It is proved mathematically that the number of computation steps required for the proposed fast matching method is less than that needed by conventional matching.

Fast word detection in a speech using new high speed time delay neural networks

Abstract: This paper presents a new approach to speed up the operation of time delay neural networks for fast detecting a word in a speech. The entire data are collected together in a long vector and then tested as a one input pattern. The proposed fast time delay neural networks (FTDNNs) use cross correlation in the frequency domain between the tested data and the input weights of neural networks. It is proved mathematically and practically that the number of computation steps required for the presented time delay neural networks is less than that needed by conventional time delay neural networks (CTDNNs). Simulation results using MATLAB confirm the theoretical computations.

An optimal robust digital image watermarking based on genetic algorithms in multiwavelet domain

Abstract: In this paper, we propose digital image watermarking algorithm in the multiwavelet transform domain. The embedding technique is based on the quantization index modulation technique and this technique does not require the original image in the watermark extraction. We have developed an optimization technique using the genetic algorithms to search for optimal quantization steps to improve the quality of watermarked image and robustness of the watermark. In addition, we analyze the performance of the proposed algorithm in terms of peak signal to noise ratio and normalized correlation. The experimental results show that our proposed method can improve the quality of the watermarked image and give more robustness of the watermark as compared to previous works.

Non-linear image representation based on IDP with NN

Abstract: In this paper is offered a method for non-linear still image representation based on pyramidal decomposition with a neural network. This approach is developed by analogy with the hypothesis for the way humans do image recognition using consecutive approximations with increasing similarity. A hierarchical decomposition, named Inverse Difference Pyramid (IDP), is used for the image representation. The approximations in the consecutive decomposition layers are represented by the neurons in the hidden layers of the neural networks (NN). This approach ensures efficient description of the processed images and as a result -- a high compression ratio. This new way for image representation is suitable for various applications (efficient compression, multi-layer search in image databases, etc.).

A novel watermarking scheme for JPEG images

Abstract: Image watermarking with both insensible detection and high robustness capabilities is still a challenging problem. Even if some of the watermarking areas involve huge financial implication, there are relatively fewer efforts presented, which primarily focus the sustainability against some attacks, which are specific to financial application area particularly. One of such application area is "Fingerprinting" and a major threat for this area is "Collusion Attack". This paper presents an inherently collusion attack resistant (ICAR) scheme for hiding a logo-based watermark in JPEG images. This scheme is based on averaging of low and middle frequency coefficients of block Discrete Cosine Transform (DCT) coefficients of an image. Experimental results show the robustness of the proposed scheme against the JPEG compression and other common image manipulations.

Real-time detection of face and iris

Abstract: In this study, a computational algorithm has been developed to automatically detect human face and irises from color images captured by real-time camera. Haar cascade-based algorithm has been applied for simple and fast face detection. The face image is then converted into grayscale image. Three types of image processing techniques have been tested respectively to study its effect on the performance of iris detection algorithm. Then, iris candidates are extracted from the valley created at the face region. The iris candidates are paired up and the cost of each possible pairing is computed by a combination of mathematical models. Finally, the positions of the detected irises are used as a reference to refine the face region. The algorithm has been tested by quality images from Logitech camera and noisy images from Voxx CCD camera. The proposed algorithm has achieved 83.60% as the highest success rate of iris detection under a user-friendly and unconstraint office environment.

Biologically inspired evolutionary computing tools for the extraction of fetal electrocardiogram

Abstract: Signals recorded from the human body provide valuable information about the biological activities of body organs. The spectral properties of different organs help in medical diagnosis. Even small changes in functioning of organs is indicated by the changes in their spectra. Fetal heart rate extraction from the abdominal ECG is of great importance because the information carried by it is helpful in assessing appropriately the fetus well-being during pregnancy. Fetal ECG is always contaminated by a drift and interference caused by several bioelectric phenomena, or by various types of noise, such as intrinsic noise from the recorder and noise from electrode-skin contact. The low Signal to noise Ratio of fetal ECG makes it difficult to analyze it effectively. Accurate detection of QRS complex is a pre-requisite in the assessment of fetal heart beat. In this paper we utilize an intelligent Adaptive Filter for noise cancellation in the effective extraction and analysis of fetal ECG. The PSO based adaptive noise cancellation technique is shown to be superior to the conventional FIR adaptive filtering.

GSM/GPRS signal strength measurements in aircraft flights under 3,000 meters of altitude

Abstract: Nowadays an increasing demand to use mobile telephones devices in aircraft flights is being acknowledged, both in commercial and in aviation general flights. 2G and 3G mobile communications networks have a great penetration in terrestrial surface of populated areas. Nevertheless land mobile networks have not been planned to operate within the air space. The main objective of this project has been to collect GSM/GPRS signal level measurement samples in the air space used by general aviation, transmitted by terrestrial base stations. Subsequently the values of obtained signal have been analyzed in order to extract conclusions on the applicability of the current mobile terrestrial communications on board of aircraft in general aviation.

The use of wavelets in speaker feature tracking identification system using neural network

Abstract: Continuous and Discrete Wavelet Transform (WT) are used to create text-dependent robust to noise speaker recognition system. In this paper we investigate the accuracy of identification the speaker identity in non- stationary signals. Three methods are used to extract the essential speaker features based on Continuous, Discrete Wavelet Transform and Power Spectrum Density (PSD). To have better identification rate, two types of Neural Networks (NNT) are studied: The first is Feed Forward Back Propagation Neural Network (FFBNN) and the second is perceptron. Up to 98.44% identification rate is achieved. The presented system depends on the multi-stage features extracting due to its better accuracy. The multistage features tracking based system shows good capability of features tracking for tested signals with SNR equals to -9 dB using Wavelet Transform, which is suitable for non-stationary signal.

Eight-phase sequence sets design for radar

Abstract: In this paper a novel Modified Simulated Annealing Algorithm (MSAA) is used as global optimization technique to find the solution of combinatorial optimization problem which is usually difficult to tackle. MSAA combines good methodologies like global minimum converging property of Simulated Annealing algorithm and fast convergence rate of Hamming scan algorithm. Orthogonal Netted Radar System (ONRS) and spread spectrum communication system can fundamentally improve the system performance by using a group of specially designed orthogonal signals. MSAA is used to synthesize orthogonal eight-phase sequence sets with good autocorrelation and cross correlation properties. Some of the synthesized sequence sets are presented, and their properties are better than four-phase sequence sets known in the literature. The synthesized eight-phase sequence sets are promising for practical application to Netted Radar System and spread spectrum communication. The effect of Doppler shift on synthesized sequences set is also investigated using ambiguity function. The convergence rate of the algorithm is shown to be good.

A novel blind digital watermarking technique for Stegano-encrypting information using nine-ac-coefficient prediction algorithm with an innovative security strategy

Abstract: This paper presents a new methodology for data hiding using digital watermarking in the DCT Domain. The methodology relies on a new scheme for encrypting the data prior to the embedding stage. The key used for ciphering is almost of arbitrary length, type and format; this endows the watermark with a powerful, 3-level reinforced security structure. It is a blind-detector watermarking technique and the amount of the hidden data is increased by 60% compared with the traditional AC-Coefficients Prediction algorithm while sustaining a high level of transparency. Simulation results were carried out which demonstrated a promising PSNR, limited blocking artifacts, and a satisfactory level of the overall performance. The paper also presents an extensive survey of prominent digital-watermarking research outcomes in the WSEAS Transactions.

An overview of different wideband direction of arrival (DOA) estimation methods

Abstract: The Direction of Arrival (DOA) estimation methods are useful in Sonar, Radar and Advanced Satellite and Cellular Communication systems. In this paper different Direction of Arrival(DOA) methods such as Coherent Signal Subspace Processing (CSSM), the Weighted Average of Signal Subspaces (WAVES) and Test of Orthogonality of Projected Subspaces (TOPS) and Incoherent MUSIC(IMUSIC) is presented and their performance is also compared. The TOPS method performs better than others in mid signal--to-noise ratio (SNR) ranges, while CSSM and WAVES work better in low SNR. Incoherent methods like IMUSIC works best at high SNR.

A cumulant-based method for gait identification using accelerometer data with principal component analysis and support vector machine

Abstract: In this paper a cumulant-based method for identification of gait using accelerometer data is presented. Acceleration data of three different walking speeds (slow, normal and fast) for each subject ...

Color video segmentation using Fuzzy C-mean clustering with spatial information

Abstract: Video segmentation can be considered as a clustering process that classifies one video succession into several objects. Spatial information enhances the quality of clustering which is not utilized in the conventional FCM. Normally fuzzy c-mean (FCM) algorithm is not used for color video segmentation and it is not robust against noise. In this paper, we presented a modified version of fuzzy c-means (FCM) algorithm that incorporates spatial information into the membership function for clustering of color videos. We used HSV model for decomposition of color video and then FCM is applied separately on each component of HSV model. For optimal clustering, grayscale image is used. Additionally, spatial information is incorporated in each model separately. The spatial function is the summation of the membership function in the neighborhood of each pixel under consideration. The advantages of this new method are: (a) it yields regions more homogeneous than those of other methods for color videos; (b) it reduces the spurious blobs; and (c) it removes noisy spots. It is less sensitive to noise as compared with other techniques. This technique is a powerful method for noisy color video segmentation and works for both single and multiple-feature data with spatial information.

Methods of measure and analyse of video quality of the image

Abstract: This paper present an analysis of the method for the evaluation of quality of the images from the video signal. The results are based on the simulation of the human perception. The device used for the evaluation is a TEKTRONIX PQA500 Picture Quality Analyzer.

An efficient QR-based selection criterion for selecting an optimal precoding matrix employed in a simplistic MIMO detection

Abstract: In multiple-input multiple-output (MIMO) channel (H) communication, when channel status information (CSI) is known to the receiver but not to the transmitter, the precoding technique can achieve a highly reliable communication link, when the receiver informs an optimal precoding matrix index to the transmitter based on current CSI. To select an optimal precoding matrix (F), the maximum capacity selection criterion and the maximum minimum singular value selection criterion are developed. However, with QR-decomposition detection (HF = QR) in the precoding system, these two selection criteria may involve high complexity and poor detection performance due to the full matrix multiplication and inaccurate detection of the first layer, respectively. In this paper, to simplify the QR-decomposition processes, the real and imaginary parts of channel elements are rearranged to achieve a column-wise orthogonal structure to reduce the repeated computation. In precoding systems, to achieve 1) low-complexity and 2) performance enhancement, the efficient QR-based selection (QR-selection) criterion is proposed to select an optimal precoding matrix by maximizing the absolute value of the lowest layer of the upper triangular matrix R. For low-complexity, to reduce the multiplication complexity of computing R, we prove that the absolute value of R is equal to the absolute value of R (RF=QR), where H = QR. Based on this equivalence, we can reduce the multiplication complexity because the number of multiplications for computing RF is less than the number of multiplications for computing HF. For performance enhancement, the proposed QR-selection criterion can effectively mitigate the impact of error-propagation because the probability of an early error in the sequence of decisions is lower. Simulation results show that the proposed scheme with a low-complexity level has a better performance than others, and that it can improve detection performance as the codebook size increases.

A simple algorithm for automated skin lesion border detection

Abstract: Prompt diagnosis is the most reliable solution for an effective treatment of melanoma. There is an ongoing research for providing computer-aided imaging tools in order to support the early detection and diagnosis of malignant melanomas. The first step towards producing such a diagnosis system is the automated and accurate boundary detection of skin lesion. Therefore, the present study introduces a new, simple, and very fast algorithm that has the ability to detect effectively and automatically the border of potential melanoma. The complexity of the proposed algorithm is O(√N), and thus the execution time, is dramatically minimized.

Signal processing's importance in manufacturing of a special device

Abstract: Lot of the parts manufactured for various required purposes involve machining processes, such as turning, drilling or, cold pressing processes, such as stamping, drawing, extruding, etc. When these processes are involved, special attention should be given to their specific force values and, as consequence, to devices used in measuring force's values. The most important element of a force measuring device is represented by elastic element and, further, by the transducers "fitted" to it. Appropriate transducers signal processing is essential when device's characteristics have to de specified and, more, to be tested and applied into real manufacturing conditions.

Simple and powerful instrument model for the source separation of polyphonic music

Abstract: This article presents a new approach to sound source separation. The introduced algorithm is based on spectral modeling of real instruments. The separation of independent sources is carried out by dividing the energy of the mixture signal based on these instrument models. This way it is possible to regain some of the information that was lost when the independent sources were mixed together into a single signal. The paper presents the theory behind the proposed separation system, then focuses on the instrument model that is the basic element of the approach. Measurement results are given for polyphony levels from 2 to 10 demonstrating the separation quality, with special regard to the effect of prints on the result.

Fast algorithms with low complexity for adaptive filtering

Abstract: The numerically stable version of fast recursive least squares (NS-FRLS) algorithms represent a very important load of calculation that needs to be reduced. Its computational complexity is of 8L operations per sample, where L is the finite impulse response filter length. We propose an algorithm for adaptive filtering, while maintaining equilibrium between its reduced computational complexity and its adaptive performances. We present a new (M-SMFTF) algorithm for adaptive filtering with fast convergence and low complexity. It is the result of a simplified FTF type algorithm, where the adaptation gain is obtained only from the forward prediction variables and using a new recursive method to compute the likelihood variable. This algorithm presents a certain interest, for the adaptation of very long filters, like those used in the problems of echo acoustic cancellation, due to its reduced complexity, its numerical stability and its convergence in the presence of the speech signal. Its computational complexity is of 6L and this is considerably reduced to 2L+4P when we use a reduced P-size (P≪L) forward predictor.

Range migration compensation based on range-direction coupling in SFDLFM MIMO radar

Abstract: Wide low-gain transmitting beam and long time integration are adopted in the orthogonal signal MIMO radar to survey the interested area. The range migration of moving target is a pivotal problem faced in the MIMO radar. Orthogonal LFM signal is one of the most familiar waveforms in MIMO radar, and this paper discusses the range migration compensation problem in the MIMO radar using Stepped Frequency Division Linear Frequency Modulation (SFDLFM) signal. A new compensation method based on the proper range- direction coupling relationship is put forward. It can achieve a good compensation effect with low computation complexity. Theoretical deduction and simulation results demonstrate the validity of this method.

Adaptive thresholding of DFT coefficients based on probability distribution of additive noise

Abstract: The proposed method of adaptive thresholding uses probability distribution of additive noise signal, by which the input signal is corrupted. The additive noise with non-uniformly distributed power spectral density can be reduced via normalization process. The method is focused on musical signal corrupted by the noise with relative high input signal-to-noise ratio ranging between 20 and 30 dB. The method uses the thresholding of coefficients of Discrete Fourier transform (DFT). Minimal signal distortion should be introduced by this method. In conclusion the method is tested for noise reduction efficiency and size of degradation of processed signal.