Abstract: Currently, extensive signal analysis is performed in order to evaluate structural health of turbomachinery blades. This approach is affected by constraints of time and the availability of qualified personnel. Thus, new approaches to blade dynamics identification that provide faster and more accurate results are sought after. Generally, modal analysis is employed in acquiring dynamic properties of a vibrating turbomachinery blade and is widely adopted in condition monitoring of blades. The analysis provides useful information on the different modes of vibration and natural frequencies by exploring different shapes that can be taken up during vibration since all mode shapes have their corresponding natural frequencies. Experimental modal testing and finite element analysis are the traditional methods used to evaluate mode shapes with limited application to real live scenario to facilitate a robust condition monitoring scheme. For a real time mode shape evaluation, rapid evaluation and low computational cost is required and traditional techniques are unsuitable. In this study, artificial neural network is developed to evaluate the mode shape of a lab scale rotating blade assembly by using result from finite element modal analysis as training data. The network performance evaluation shows that artificial neural network (ANN) is capable of mapping the correlation between natural frequencies and mode shapes. This is achieved without the need of extensive signal analysis. The approach offers advantage from the perspective that the network is able to classify mode shapes and can be employed in real time including simplicity in implementation and accuracy of the prediction. The work paves the way for further development of robust condition monitoring system that incorporates real time mode shape evaluation.
Abstract: In this work, a training algorithm for probabilistic neural networks (PNN) is presented. The algorithm addresses one of the major drawbacks of PNN, which is the size of the hidden layer in the network. By using a cross-validation training algorithm, the number of hidden neurons is shrunk to a smaller number consisting of the most representative samples of the training set. This is done without affecting the overall architecture of the network. Performance of the network is compared against performance of standard PNN for different databases from the UCI database repository. Results show an important gain in network size and performance.
Abstract: The issue of high blood sugar level, the effects of which might end up as diabetes mellitus, is now becoming a rampant cardiovascular disorder in our community. In recent times, a lack of awareness among most people makes this disease a silent killer. The situation calls for urgency, hence the need to design a device that serves as a monitoring tool such as a wrist watch to give an alert of the danger a head of time to those living with high blood glucose, as well as to introduce a mechanism for checks and balances. The neural network architecture assumed 8-15-10 configuration with eight neurons at the input stage including a bias, 15 neurons at the hidden layer at the processing stage, and 10 neurons at the output stage indicating likely symptoms cases. The inputs are formed using the exclusive OR (XOR), with the expectation of getting an XOR output as the threshold value for diabetic symptom cases. The neural algorithm is coded in Java language with 1000 epoch runs to bring the errors into the barest minimum. The internal circuitry of the device comprises the compatible hardware requirement that matches the nature of each of the input neurons. The light emitting diodes (LED) of red, green, and yellow colors are used as the output for the neural network to show pattern recognition for severe cases, pre-hypertensive cases and normal without the traces of diabetes mellitus. The research concluded that neural network is an efficient Accu-Chek design tool for the proper monitoring of high glucose levels than the conventional methods of carrying out blood test.
Abstract: The use of decision support systems in agriculture may help monitoring large fields of crops by automatically detecting the symptoms of foliage diseases. In our work, we designed and implemented a decision support system for small tomatoes producers. This work investigates ways to recognize the late blight disease from the analysis of digital images of tomatoes, using a pair of multilayer perceptron neural networks. The networks outputs are used to generate repainted tomato images in which the injuries on the plant are highlighted, and to calculate the damage level of each plant. Those levels are then used to construct a situation map of a farm where a cellular automata simulates the outbreak evolution over the fields. The simulator can test different pesticides actions, helping in the decision on when to start the spraying and in the analysis of losses and gains of each choice of action.
Abstract: Image rotation is one of main pre-processing steps for image processing or image pattern recognition. It is implemented with a rotation matrix multiplication. It requires a lot of floating point arithmetic operations and trigonometric calculations, so it takes a long time to execute. Therefore, there has been a need for a high speed image rotation algorithm without two major time-consuming operations. However, the rotated image has a drawback, i.e. distortions. We solved the problem using an augmented two-step shear transform. We compare the presented algorithm with the conventional rotation with images of various sizes. Experimental results show that the presented algorithm is superior to the conventional rotation one.
Abstract: Foliage diseases in plants can cause a reduction in both quality and quantity of agricultural production. Intelligent detection of plant diseases is an essential research topic as it may help monitoring large fields of crops by automatically detecting the symptoms of foliage diseases. This work investigates ways to recognize the late blight disease from the analysis of tomato digital images, collected directly from the field. A pair of multilayer perceptron neural network analyzes the digital images, using data from both RGB and HSL color models, and classifies each image pixel. One neural network is responsible for the identification of healthy regions of the tomato leaf, while the other identifies the injured regions. The outputs of both networks are combined to generate the final classification of each pixel from the image and the pixel classes are used to repaint the original tomato images by using a color representation that highlights the injuries on the plant. The new images will have only green, red or black pixels, if they came from healthy or injured portions of the leaf, or from the background of the image, respectively. The system presented an accuracy of 97% in detection and estimation of the level of damage on the tomato leaves caused by late blight.
Abstract: The beginning of 21st century has witnessed new
advancements in the design and use of new materials for biosensing
applications, from nano to macro, protein to tissue. Traditional
analytical methods lack a complete toolset to describe the
complexities introduced by living systems, pathological relations,
discrete hierarchical materials, cross-phase interactions, and
structure-property dependencies. Materiomics – via systematic
molecular dynamics (MD) simulation – can provide structureprocess-
property relations by using a materials science approach
linking mechanisms across scales and enables oriented biosensor
design. With this approach, DNA biosensors can be utilized to detect
disease biomarkers present in individuals’ breath such as acetone for
diabetes. Our wireless sensor array based on single-stranded DNA
(ssDNA)-decorated single-walled carbon nanotubes (SWNT) has
successfully detected trace amount of various chemicals in vapor
differentiated by pattern recognition. Here, we present how MD
simulation can revolutionize the way of design and screening of DNA
aptamers for targeting biomarkers related to oral diseases and oral
health monitoring. It demonstrates great potential to be utilized to
build a library of DNDA sequences for reliable detection of several
biomarkers of one specific disease, and as well provides a new
methodology of creating, designing, and applying of biosensors.
Abstract: Due to the fact that there exist only a small number of complex systems in artificial immune system (AIS) that work out nonlinear problems, nonlinear AIS approaches, among the well-known solution techniques, need to be developed. Gaussian function is usually used as similarity estimation in classification problems and pattern recognition. In this study, diagnosis of breast cancer, the second type of the most widespread cancer in women, was performed with different distance calculation functions that euclidean, gaussian and gaussian-euclidean hybrid function in the clonal selection model of classical AIS on Wisconsin Breast Cancer Dataset (WBCD), which was taken from the University of California, Irvine Machine-Learning Repository. We used 3-fold cross validation method to train and test the dataset. According to the results, the maximum test classification accuracy was reported as 97.35% by using of gaussian-euclidean hybrid function for fold-3. Also, mean of test classification accuracies for all of functions were obtained as 94.78%, 94.45% and 95.31% with use of euclidean, gaussian and gaussian-euclidean, respectively. With these results, gaussian-euclidean hybrid function seems to be a potential distance calculation method, and it may be considered as an alternative distance calculation method for hard nonlinear classification problems.
Abstract: Work is in on line Arabic character recognition and the principal motivation is to study the Arab manuscript with on line technology.
This system is a Markovian system, which one can see as like a Dynamic Bayesian Network (DBN). One of the major interests of these systems resides in the complete models training (topology and parameters) starting from training data.
Our approach is based on the dynamic Bayesian Networks formalism. The DBNs theory is a Bayesians networks generalization to the dynamic processes. Among our objective, amounts finding better parameters, which represent the links (dependences) between dynamic network variables.
In applications in pattern recognition, one will carry out the fixing of the structure, which obliges us to admit some strong assumptions (for example independence between some variables). Our application will relate to the Arabic isolated characters on line recognition using our laboratory database: NOUN. A neural tester proposed for DBN external optimization.
The DBN scores and DBN mixed are respectively 70.24% and 62.50%, which lets predict their further development; other approaches taking account time were considered and implemented until obtaining a significant recognition rate 94.79%.
Abstract: In pattern clustering, nearest neighborhood point computation is a challenging issue for many applications in the area of research such as Remote Sensing, Computer Vision, Pattern Recognition and Statistical Imaging. Nearest neighborhood
computation is an essential computation for providing sufficient classification among the volume of pixels (voxels) in order to localize the active-region-of-interests (AROI). Furthermore, it is needed to compute spatial metric relationships of diverse area of imaging based on the applications of pattern recognition. In this paper, we propose a new methodology for finding the nearest neighbor point, depending on making a virtually grid of a hexagon cells, then locate every point beneath them. An algorithm is suggested for minimizing the computation and increasing the turnaround time of the process. The nearest neighbor query points Φ are fetched by seeking fashion of hexagon holistic. Seeking will be repeated until an AROI Φ is to be expected. If any point Υ is located then searching starts in the nearest hexagons in a circular way. The First hexagon is considered be level 0 (L0) and the surrounded hexagons is level 1 (L1). If Υ is located in L1, then search starts in the next level (L2) to ensure that Υ is the nearest neighbor for Φ. Based on the result and experimental results, we found that the proposed method has an advantage over the traditional methods in terms of minimizing the time complexity required for searching the neighbors, in turn, efficiency of classification will be improved sufficiently.
Abstract: Learning the gradient of neuron's activity function
like the weight of links causes a new specification which is
flexibility. In flexible neural networks because of supervising and
controlling the operation of neurons, all the burden of the learning is
not dedicated to the weight of links, therefore in each period of
learning of each neuron, in fact the gradient of their activity function,
cooperate in order to achieve the goal of learning thus the number of
learning will be decreased considerably.
Furthermore, learning neurons parameters immunes them against
changing in their inputs and factors which cause such changing.
Likewise initial selecting of weights, type of activity function,
selecting the initial gradient of activity function and selecting a fixed
amount which is multiplied by gradient of error to calculate the
weight changes and gradient of activity function, has a direct affect
in convergence of network for learning.
Abstract: Writer identification is one of the areas in pattern
recognition that attract many researchers to work in, particularly in
forensic and biometric application, where the writing style can be
used as biometric features for authenticating an identity. The
challenging task in writer identification is the extraction of unique
features, in which the individualistic of such handwriting styles
can be adopted into bio-inspired generalized global shape for
writer identification. In this paper, the feasibility of generalized
global shape concept of complimentary binding in Artificial
Immune System (AIS) for writer identification is explored. An
experiment based on the proposed framework has been conducted
to proof the validity and feasibility of the proposed approach for
off-line writer identification.
Abstract: Recognizing human action from videos is an active
field of research in computer vision and pattern recognition. Human
activity recognition has many potential applications such as video
surveillance, human machine interaction, sport videos retrieval and
robot navigation. Actually, local descriptors and bag of visuals words
models achieve state-of-the-art performance for human action
recognition. The main challenge in features description is how to
represent efficiently the local motion information. Most of the
previous works focus on the extension of 2D local descriptors on 3D
ones to describe local information around every interest point. In this
paper, we propose a new spatio-temporal descriptor based on a spacetime
description of moving points. Our description is focused on an
Accordion representation of video which is well-suited to recognize
human action from 2D local descriptors without the need to 3D
extensions. We use the bag of words approach to represent videos.
We quantify 2D local descriptor describing both temporal and spatial
features with a good compromise between computational complexity
and action recognition rates. We have reached impressive results on
publicly available action data set
Abstract: Persian (Farsi) script is totally cursive and each character is written in several different forms depending on its former and later characters in the word. These complexities make automatic handwriting recognition of Persian a very hard problem and there are few contributions trying to work it out. This paper presents a novel practical approach to online recognition of Persian handwriting which is based on representation of inputs and patterns with very simple visual features and comparison of these simple terms. This recognition approach is tested over a set of Persian words and the results have been quite acceptable when the possible words where unknown and they were almost all correct in cases that the words where chosen from a prespecified list.
Abstract: Automatic currency note recognition invariably
depends on the currency note characteristics of a particular country
and the extraction of features directly affects the recognition ability.
Sri Lanka has not been involved in any kind of research or
implementation of this kind. The proposed system “SLCRec" comes
up with a solution focusing on minimizing false rejection of notes.
Sri Lankan currency notes undergo severe changes in image quality
in usage. Hence a special linear transformation function is adapted to
wipe out noise patterns from backgrounds without affecting the
notes- characteristic images and re-appear images of interest. The
transformation maps the original gray scale range into a smaller
range of 0 to 125. Applying Edge detection after the transformation
provided better robustness for noise and fair representation of edges
for new and old damaged notes. A three layer back propagation
neural network is presented with the number of edges detected in row
order of the notes and classification is accepted in four classes of
interest which are 100, 500, 1000 and 2000 rupee notes. The
experiments showed good classification results and proved that the
proposed methodology has the capability of separating classes
properly in varying image conditions.
Abstract: Fuzzy C-means Clustering algorithm (FCM) is a
method that is frequently used in pattern recognition. It has the
advantage of giving good modeling results in many cases, although,
it is not capable of specifying the number of clusters by itself. In
FCM algorithm most researchers fix weighting exponent (m) to a
conventional value of 2 which might not be the appropriate for all
applications. Consequently, the main objective of this paper is to use
the subtractive clustering algorithm to provide the optimal number of
clusters needed by FCM algorithm by optimizing the parameters of
the subtractive clustering algorithm by an iterative search approach
and then to find an optimal weighting exponent (m) for the FCM
algorithm. In order to get an optimal number of clusters, the iterative
search approach is used to find the optimal single-output Sugenotype
Fuzzy Inference System (FIS) model by optimizing the
parameters of the subtractive clustering algorithm that give minimum
least square error between the actual data and the Sugeno fuzzy
model. Once the number of clusters is optimized, then two
approaches are proposed to optimize the weighting exponent (m) in
the FCM algorithm, namely, the iterative search approach and the
genetic algorithms. The above mentioned approach is tested on the
generated data from the original function and optimal fuzzy models
are obtained with minimum error between the real data and the
obtained fuzzy models.
Abstract: Random and natural textures classification is still
one of the biggest challenges in the field of image processing and
pattern recognition. In this paper, texture feature extraction using
Slant Hadamard Transform was studied and compared to other
signal processing-based texture classification schemes. A
parametric SHT was also introduced and employed for natural
textures feature extraction. We showed that a subtly modified
parametric SHT can outperform ordinary Walsh-Hadamard
transform and discrete cosine transform. Experiments were carried
out on a subset of Vistex random natural texture images using a
kNN classifier.
Abstract: The Ant Colony Optimization (ACO) is a metaheuristic inspired by the behavior of real ants in their search for the shortest paths to food sources. It has recently attracted a lot of attention and has been successfully applied to a number of different optimization problems. Due to the importance of the feature selection problem and the potential of ACO, this paper presents a novel method that utilizes the ACO algorithm to implement a feature subset search procedure. Initial results obtained using the classification of speech segments are very promising.
Abstract: This paper presents and evaluates a new classification
method that aims to improve classifiers performances and speed up
their training process. The proposed approach, called labeled
classification, seeks to improve convergence of the BP (Back
propagation) algorithm through the addition of an extra feature
(labels) to all training examples. To classify every new example, tests
will be carried out each label. The simplicity of implementation is the
main advantage of this approach because no modifications are
required in the training algorithms. Therefore, it can be used with
others techniques of acceleration and stabilization. In this work, two
models of the labeled classification are proposed: the LMLP
(Labeled Multi Layered Perceptron) and the LNFC (Labeled Neuro
Fuzzy Classifier). These models are tested using Iris, wine, texture
and human thigh databases to evaluate their performances.
Abstract: Everyday the usages of the Internet increase and simply a world of the data become accessible. Network providers do not want to let the provided services to be used in harmful or terrorist affairs, so they used a variety of methods to protect the special regions from the harmful data. One of the most important methods is supposed to be the firewall. Firewall stops the transfer of such packets through several ways, but in some cases they do not use firewall because of its blind packet stopping, high process power needed and expensive prices. Here we have proposed a method to find a discriminate function to distinguish between usual packets and harmful ones by the statistical processing on the network router logs. So an administrator can alarm to the user. This method is very fast and can be used simply in adjacent with the Internet routers.