International Journal of Image, Graphics and Signal Processing @ijigsp
Статьи журнала - International Journal of Image, Graphics and Signal Processing
Все статьи: 1110

Статья научная
In synthetic aperture radar (SAR) imaging system speckle is modeled as a multiplicative noise which limits the performance of SAR image processing systems. In the literature, several SAR image despeckling algorithms have been presented, among them two simple, yet effective, approaches are using thresholding and Bayesian estimation in transform domains. In this article, we try to provide proper answer to this question: which one of these two despeckling methods works better? To this aim, we first introduce a new thresholding function with two thresholds, and show that when thresholds are determined through optimization procedures, an improved denoising performance in terms of joint speckle removal and edge saving efficiencies can be achieved. However, still a Bayesian LMMSE/MAP estimator can provide greater speckle removal efficiency in test images with high speckle power, and some thresholding methods produce better edge saving efficiency. Hence, aiming at joint exploitation of the superior edge saving ability of thresholding estimator and greater speckle removal efficiency of Bayesian estimator, we next propose the idea of using a combined despecking algorithm. The new denoising methods are applied for despeckling of true SAR images in the nonsubsampled contourlet transform domain and the situations they achieve superior performance have been highlighted.
Бесплатно

Time-varying Comb Filters to Improve Speech Perception in Sensorineural Hearing Loss Subjects
Статья научная
In the case of Sensorineural Hearing Loss (SNHL) persons speech perception diminishes in a noisy environment because of masking. The present work aims mainly at improving speech perception in sensorineural hearing-impaired subjects, as there is no known medical treatment for this condition. Speech perception can be improved by reducing the impact of masking. This is accomplished by splitting the speech signal into two parts for binaural dichotic presentation using time-varying comb filters having complementary magnitude responses. Using the frequency sampling method time-varying comb (FIR) filters with magnitude responses complementary to each other with 512 order are designed to split the speech signal for dichotic presentation. For the purpose of designing filters, 22 kHz sampling frequency and twenty-two one-third octave bands spanning from 0 to 11 kHz are taken into consideration. Magnitude responses of filters are continuously swept with a time shift less than just noticeable difference (JND) so that capacity to detect gaps in speech signal enhances without negating the benefits of the spectral splitting technique. Filter functioning is evaluated by using objective and subjective measures. Using Perceptual Evaluation of Speech Quality (PESQ) and spectrographic analysis an objective evaluation is made. The subjective measure is done using Mean Opinion Score (MOS) for quality of speech. MOS test is examined on normal hearing subjects by adding white noise to study materials at different SNR levels. For the evaluation of intelligibility of speech Modified Rhyme Test (MRT) is considered and evaluated on normal hearing subjects as well as bilateral moderate SNHL persons by adding white noise to study materials at different SNR levels. Study materials used for the evaluation of quality are VC syllable /aa-b/ & vowel /aa/. 300 monosyllabic words of consonant-vowel-consonant (CVC) are used as study materials for the evaluation of speech intelligibility. The outcomes showed an improvement in PESQ values and MOS test scores for lower SNR values comparing unprocessed speech with processed speech and also an improvement in the intelligibility of processed speech in a noisy atmosphere for both types of subjects. Thus there is an enhancement in speech perception of processed speech in a noisy environment.
Бесплатно

Tomographic Convex Time-Frequency Analysis
Статья научная
In this paper we aim to solve a problem of image reconstruction in tomography. In medical imaging, patients suffer from taking high dose of radioactive drug in order to get a well-qualified image. Our goal is to reduce this dose of radioactive drug given to the patients in PET scan and to get a well-qualified image. We use to modeling this problem using a convex function to minimize. In tomography, real problem requires a positive constraint and may get a blurred image due to poisson noise. Then, in order to get back a non blurred image of human body, we add to this function a wavelet regularization which is a non differentiable function. We introduce specific algorithms to get the minimum of the global function obtained. After presenting the classic algorithms with their conditions to solve the problem we find that Chambolle Pock's algorithm requires less properties than these algorithms and gives good results. Then, we propose its computation method with the proof.
Бесплатно

Towards Query Efficient and Derivative Free Black Box Adversarial Machine Learning Attack
Статья научная
While deep learning has shown phenomenal success in many critical applications such as in autonomous driving and medical diagnosis, it is vulnerable to black box adversarial machine learning attacks. Objective of these attacks is to mislead a classifier in making mistakes. Hard Label attacks are those in which an adversary has access only to the top-1 prediction label and has no knowledge about model parameters or gradient loss. Secondly, for security concerns, the number of model queries that an attacker can perform for evaluation are restricted. In this paper, we propose a novel nature-inspired optimization algorithm for generating adversarial examples. Proposed algorithm is derivative-free, meta-heuristic algorithm. It searches for optimum adversarial examples in high-dimensional image space using simple arithmetic operations inspired by Brownian motion of molecules in fluids and gases. Experiments with CIFAR-10 image dataset yielded encouraging results with a query budget of less than 1000 and with a minimal distortion to original image. Its performance was determined to be comparable and exceeded in some cases compared to previous state of the art attacks.
Бесплатно

Статья научная
An electrocardiogram (ECG) machine is a device that checks the patient’s heart rhythm and electrical activity. This is done by attaching sensors on the skin of the patients. But the problem with these machines is that, these are expensive and not portable. Thus it is difficult to use these machines in the rural or remote areas of developing countries like Bangladesh where the issue of portability and cost arises. In this paper, the problem of cost and portability is addressed. We propose a complete solution for a low-cost portable ECG monitoring from recording to report generation for patients including real-time ECG traces on screen with storage options and calculation of all necessary diagnostics parameters for helping the doctors to make decision. This type of ECG machines could be used in hospitals, homes, villages or even in a disaster area. The system designed in this paper includes a PC/Laptop, as these devices are now widely available, at least available at hospitals and health care centers, even in rural/remote areas of Bangladesh. With this, a significant difference can be made against heart diseases.
Бесплатно

Tracking of Moving Object Using Centroid based Prediction and Boundary Tracing Scheme
Статья научная
Object tracking has always been a hotspot in the field of computer vision and has myriad applications in the real world. A major problem in this field is that of the successful tracking of a moving object undergoing occlusion in its path. This paper presents centroid based tracking scheme of a moving object without any apriori information of its shape or motion. Once the boundary of the object of interest is obtained, the centroid is calculated from its first order moments. This centroid is further utilized to detect the partial occlusion of test object by some other still or moving object in image frame. In case occlusion is detected, the new centroid location of moving object is predicted for subsequent video frames. The proposed algorithm is able to successfully detect moving object undergoing partial or total occlusion. Experimental results of our algorithm are compared with a popular tracking technique based on Mean Shift tracking algorithm.
Бесплатно

Traffic video enhancement based vehicle correct tracked methodology
Статья научная
In this paper, an enhancement based traffic video has been proposed in the state of the art of computer vision. The main target is to develop a decision making criteria for removing the most probable video degradations. Such traffic video degradations would have an adverse impact on the transportation system. In order to establish the appropriate analysis, three types of video degradations have been added to the test video; salt and pepper noise, Gaussian noise, and speckle noise, we have simulated rainy, fog, and darkness conditions for the traffic video. First of all, back ground subtraction and Kalman filter techniques have been used for detecting and tracking vehicles respectively. By using such algorithms, it would be easily to estimate average number of assigned tracks which express the efficacy of correct detection and prediction of vehicles in each frame. Furthermore, video degradations would be applied in order to studying its effect on the average number of assigned tracks which would be deviated than noiseless video. Spatial filtering system has been applied to state the most suitable filter mask which satisfy the least deviation in the average number of assigned tracks. Experimental results show that median filter satisfies the least deviation in all cases of video degradations.
Бесплатно

Transfer Subspace Learning Model for Face Recognition at a Distance
Статья научная
Many machine learning algorithms work under the assumption that the training and testing data are drawn from the same distribution. However, in practice the assumption might not hold. Transfer subspace learning algorithms aims at utilizing knowledge gained in source domain to learn a task in target domain. The main objective of this work is to apply transfer subspace learning framework on face recognition task at a distance. In this paper we identify face recognition at distance as a transfer learning problem. We show that if the face recognition task is modeled as transfer learning problem, the overall classification rate is increased significantly compared to traditional brute force approach. We also discuss a data set which is unique and meant to advance this research. The novelty of this work lies in modeling face recognition task at distance as a transfer subspace learning problem.
Бесплатно

Transformation of Classical to Quantum Image, Representation, Processing and Noise Mitigation
Статья научная
Quantum and classical computers have drastically different image representations. In a classical computer, bits are used. However, in a quantum computer, qubits are used. In this paper, the quantum image representation is the similar as the classical image representation. To represent quantum images, qubits and their associated properties have been used. Quantum imaging has previously been done via superposition. As a result, quantum imaging implemented using the superposition feature. Unitary matrices are then used to represent quantum circuits. For the quantum representation, we've gone with a modest image. To create quantum circuits, IBM's Qiskit software and Anaconda Python was used. On an IBM real time computer and an Aer simulator, a quantum circuit with 10,000 shots runs. Noise has been reduced more in the IBM real time computer than in the IBM Aer simulator. As a result, the Aer simulator's noise and qubit errors are higher than the IBM real time computer's. Quantum circuit design and image processing are both done with Qiskit programming, which is an appendix at the end of the paper. As the number of shots raise, the noise level decreases even further. Noise and qubit errors increase when the image operates at a low number of shots. Quantum image processing, noise reduction, and error correction done by circuit computation shots increase. Quantum image processing, representation, noise reduction, and error correction all make use of the quantum superposition concept.
Бесплатно

Статья научная
Realistic knowledge of rainfall characteristics and modeling parameters such as size, shape, and drop size distribution is essential in numerous areas of scientific, engineering, industrial and technological applications. Some key application areas include, but not limited to microphysics analysis of precipitation composition phenomenon, weather prediction, signal attenuations forecasting, signal processing, remote sensing, radar meteorology, stormwater management and cloud photo detection. In this contribution, the influence of rain intensity on raindrop diameter and specific attenuation in Lokoja, a typical climate region of Nigeria is investigated and reported. Three different rain rates classes obtained due to heavy rainfall depth, heavy rainfall depth, and heavy rainfall depth have been explored for the raindrop size distribution analysis. The three-parameter lognormal and Weibull models were utilised to estimate the influence of rain rates on the drop sizes and specific rainfall attenuation in the study location. For Lognormal model, the maximum raindrop concentration occurred approximately at diameter of 1 mm before showing downfall performance trends as the drop diameter increases. In the case of Weilbull model, the maximum raindrop concentration occurred at different drop diameter with the three rain rate classes, before showing downfall concentration trends with increasing rain drop diameter values. By means of the two models, the highest raindrops concentration values attained in correspondence with the specific rain attenuation were made by drop diameters not more than 2.5 mm. In terms of rain rate, specific attenuation and frequency connection, the results disclose that attenuation of propagated electromagnetic waves increases at increasing rainfall depth and increasing operating frequency bands. The results also disclose that the specific attenuation is directly proportional to the increase in rain intensity levels in correspondent with the operational frequency. As a case in point, at 4GHz frequency, the attenuation level of about 20 dB/km level is attained for mean, minimum and maximum rain rates of 29.12, 12.23 and 50.22 mm/hr, respectively. But as the frequency increased from 4GHz to 20GHz, the attenuation level almost doubles from 20 to 45dB/km at still same rain rates. The above performance is so, because at higher radio-microwave frequencies, the wavelength of the propagated electromagnetic waves approaches the mean diameter of the raindrop. The results display gradual increase in attenuation levels as the diameter rain drop sizes and intensity increases or become broader. The attenuation grows because the raindrops interfere, distort, absorb and scatter major portion of the microwave energy. However, the gradual trend in the attenuation level increase becomes slower and tending to logarithm stability at larger rain drop values. This may suggest that the attenuation level may come to equilibrium state at higher rain drop diameters. The resultant outcome of this work can assist microwaves communication engineers and relevant stakeholders in the telecommunication sector with expedient information needed to manage specific attenuation problems over Earth–space links communication channels, particualry during rainy seasons.
Бесплатно

Статья научная
Breast cancer is most common and leading cause of death among women. With improvement in the imaging modalities it is possible to diagnose the cancer at an early stage moreover treatment at an early stage reduces the mortality rate. B-mode ultrasound (US) imaging is very illustrious and reliable technique in early detection of masses in the breast. Though it is complimentary to the mammography, dense breast tissues can be examined more efficiently and detects the small nodules that are usually not observed in mammography. Segmentation of US images gives the clear understanding of nature and growth of the tumor. But some inherent artifact of US images makes this process difficult and computationally inefficient. Many methods are discussed in the literature for US image segmentation, each method has its pros and cons. In this paper, initially region merging based watershed and marker-controlled watershed transforms are discussed and implemented. In the subsequent sections we proposed a method for segmentation, based on clustering. Proposed method consists of three stages, in first stage probability images and its equalized histogram images are obtained from the original US images without any preprocessing. In the next stage, we used VQ based clustering technique with LBG, KPE and KEVR codebook generation algorithm followed by sequential cluster merging. Last stage is the post processing, where we removed unwanted regions from the selected cluster image by labeling the connected components and moreover used morphological operation for closing the holes in the final segmented image. Finally, results by our method are compared with initially discussed methods.
Бесплатно

Two-Dimensional Parameters Estimation
Статья научная
A parametric approach algorithm based on maximum likelihood estimation (MLE) method is proposed which can be exploited for high-resolution parameter estimation in the domain of signal processing applications. The array signal model turns out to be a superposition of two-dimensional sinusoids with the first component of each frequency doublet corresponding to the direction of the target and second component to the velocity. Numerical simulations are presented to illustrate the validity of the proposed algorithm and its various aspects. Also, the presented algorithm is compared with a subspace based technique, multiple signal classification (MUSIC) to highlight the key differences in performance under different circumstances. It is observed that the developed algorithm has satisfactory performance and is able to determine the direction of arrival (DOA) as well as the velocity of multiple moving targets and at the same time it performs better than MUSIC under correlated noise.
Бесплатно

Types of navigation from websites and analyzing their characteristics
Статья научная
The subject of the research in this scientific paper is a description of the opportunity for navigation through the websites, with special emphasis on analyzing different types of navigation systems. The important part of my paper is analyzing the characteristics of various navigation structures on the web pages, as well as the analyzing technical methods of displaying navigation on various users' computers. In addition, users can have different browsers, different operating systems, and different preferences in terms of their computers' settings. All these technical issues will have an impact on how the web pages will look on the user's computer. Another thing described in my work is the interpretation of navigation structures on the user's computer monitor. A special overview is made for the correlation between web navigation and all the other graphic elements in the web pages from point of view of a visual harmony of the websites. Additionally, here I give overall directions for using navigation type and their characteristics when designing websites, same as the description of some opinions and advice on the same topic. After that, I analyze twelve problems which arise from displaying navigation on the websites on the user's computer. In this paper, I will come across a few solutions for all of them, as well as recommendations for when to choose which solution.
Бесплатно

Статья научная
Following the "on condition maintenance" approach used for extending service life of an aircraft one of the major tasks is a nondestructive testing of its critical elements. Considering that many of elements of operated aircraft are manufactured from polymeric composites a special attention should be paid for diagnosing these elements due to their high vulnerability to barely visible impact damage. One of the primary testing techniques used for inspection of aircraft composite elements is an ultrasonic C-Scan technique which application results in planar images of emitted/received wave attenuation and a time of flight map. Due to the complex nature of barely visible impact damage occurrence it is difficult to analyze resulting C-Scan images. Therefore, using assistance based on image processing may help with "big–data" analysis of collected images. In this paper the authors proposed the image processing algorithm for semi-automatic evaluation of such damage distribution in aircraft composite structures. The algorithm is based on multilevel Otsu thresholding and morphological processing. Using the proposed algorithm an extraction of damage visualization from a C-Scan image as well as its characterization and 3D representation is possible. The developed approach will allow supporting diagnosing of composite structures with impact damage using C-Scan technique.
Бесплатно

Underwater Image Refinement: Color Distance and Image Formation Model (DIMFM)
Статья научная
Underwater photography is frequently used for research purpose in various domains. Domains caters to archaeology, surveillance of aquatic life movements, oceanic changes leading to alterations in weather and many more. Scientists are eager to investigate the mysterious undersea environment. For underwater surveys, archaeology departments and weather forecasting scientists obtain undersea photos. The underwater imagery however has low vision and contrast due to haze. The elimination of haze could be difficult because it depends on depth information that is unclear. Moreover, it’s challenging and complicated to clear the haze so as to enhance the underwater image. According to the investigation, fog removal algorithms currently in use do not take noise reduction approaches into account. Dehazing techniques have a hard time dealing with areas that are unevenly and excessively light. Therefore, it is vital to alter current techniques in order to make them more efficient. This work presents an innovative integrated underwater picture restoration technique. The proposed technique is in line to a pre-determined technique namely Underwater Image Formation Model. The new approach combines Bilateral Filtering, Contrast Limited Adaptive Histogram Equalization and Dark Channel Prior for better results. First, the underwater image undergoes bilateral filtering to eliminate color discrepancies. The improved image is output of the differentiation between forward and background channel. Further, the Contrast Limited Adaptive Histogram Equalizations methodology is used to produce contrast-enhanced images. Experimental results signpost that the proposed novel technique generates superior visual effects compared to other widely used undersea color image quality evaluation techniques.
Бесплатно

User Name Alias Extraction in Emails
Статья научная
Finding out user identity information from emails is one of the important research topics in email mining. Most approaches extract an email user’s name only from the header of an email, but there are often many name information appearing in the body of emails, and those names are usually more suitable for representing the sender’s or recipient’s identity. This paper focuses on the problem of extracting email users’ name aliases in the body of plain-text emails. After locating and extracting salutation and signature blocks from email bodies, we can identify the potential aliases in the salutation and signature lines, which can be directly associated with the corresponding email address in email headers, by using named entity recognition(NER) tools. However the identified aliases may be half-baked or there are still some potential aliases that can’t be correctly identified. So we propose a novel approach to efficiently and accurately extract aliases in the salutation and signature lines based on name boundary word template built on the characteristics of alias neighboring words. Results on the public subset of the Enron corpus indicate that the approaches presented in this paper can efficiently extract user’s aliases from email bodies.
Бесплатно

Статья научная
Nowadays, Karyotype analysis is frequently used in cytogenetics. It is a time-consuming and repetitive work therefore an automatic analysis can greatly be valued. In this research, an automatic method is presented. Firstly, a proposed locally adaptive thresholding method is used to segment chromosome clusters. Then, the clusters is divided into two main categories including, single chromosomes and multi-chromosome clusters based on geometric shape of clusters. In the next step, each extracted cluster is investigated to find the dark paths in order to detect touching chromosomes. Then, overlapping chromosomes are separated in clusters based on their geometric shapes. Finally, a criterion function is used to measure the similarity between the outputs of the proposed algorithm and the single chromosomes in order to recognize separated parts. The proposed algorithm is applied on 47 G-band images. The results shows that single chromosomes and clusters are recognized by the precision of 98.5% and 86.4%, respectively and separation of touching and overlapping clusters are done by precision of 70% and 67%, respectively.
Бесплатно

Utilization of Textural Features in Video Retrieval System by Hand-writing Sketch
Статья научная
In this paper we propose a video retrieval system using texture features from a hand-drawn sketch. There is currently a lot of video content on the Internet due in part to the development of video-sharing Web sites, and sometimes users are unable to narrow the options down to the desired video because there is simply too much content to choose from. We should be able to solve this problem with a retrieval technique that uses internal movie features. Previously, our laboratory proposed a video retrieval system using three internal movie features and a sketch query. In the current study, our aim is to improve the retrieval precision by adding a new feature derived from texture.
Бесплатно

Utilizing GVF Active Contours for Real-Time Object Tracking
Статья научная
In this paper an object tracking system with utilizing optical flow technique, and Gradient Vector Flow (GVF) active contours is presented. Optical flow technique is less sensitive to background structure and does not need to build a model for the background of image so it would need less time to process the image. GVF active snakes have good precision for image segmentation. However, due to the high computational cost, they are not usually applicable. Since precision and time complexity are the most important factors in the image segmentation, several methods have been developed to overcome these problems. In this paper, we, first, recognize the moving object. Then, the object fame with some pixels surrounding to it, was created. Then, this new frame is sent to the GVF filed calculation procedure. Contour initialization is obtained based on the selected pixels. This approach increases the calculation speed, and therefore makes it possible to use the contour for the tracking. The system was built, and tested with a microcomputer. The results show a speed of 10 to 12 frames per second which is considerably suitable for object tracking approaches.
Бесплатно

Variable Synergic Squeeze Convoluted Equilibrium Network Enabled Crowd Management in IoT
Статья научная
In IoT, Crowd counting is a difficult task, because of any sudden incidents people unites in a particular place. To count them effectively a crowd counting mechanism is needed. The crowd counting is help for public security. Several methods are proposed for crowd counting, but the existing methods does not provide high accuracy and high error rate. To overcome these drawbacks a Variable Synergic Squeeze Convoluted Equilibrium Network Enabled Crowd Management in IoT (VS2CEN-CC-IOT) is proposed in this manuscript for crowd counting and crowd density detection. Initially, the images are taken from two datasets named ShanghaiTech and Venice dataset. Then the images are preprocessed using Gaussian filter based preprocessing. After preprocessing the discrete wavelet transform (DWT) is used for extracting the features. The extracted features are then given to Synergic Squeeze Convoluted Equilibrium Network (SSCEN) for detecting crowd count and crowd density. In this work, variable Equilibrium Optimization Algorithm (EOA) is applied to optimize the weight parameter of SSCEN. The simulation procedure is performed in PYTHON platform. The VS^2CEN-CC-IOT attains 0.8%, 1.3%, 1.5% higher accuracy, 13%, 3.3%, 8.2% higher Precision, 12%, 10%, 17% higher specificity , 8.2%, 3.3%, 6.9% higher F1-score and 0.12%, 0.06%, 0.07% lower mean absolute error (MAE), 0.2%, 0.25%, 0.1% lower root mean square error than the existing optimization approaches such as Arithmetic Optimization Algorithm(ADA), Chaos Game Optimization(CGO) and Gradient Based Optimizer(GBO) respectively.
Бесплатно