Mohammad Mehdi Piroozmandan; Fardad Farokhi; Kaveh Kangarloo; Mohsen Jahanshahi
Articles in Press, Accepted Manuscript, Available Online from 18 August 2021
Abstract
In this paper, a novel method named RHDINTPM (Removing High Density Impulse Noise via a Novel Two Phase Method) is proposed for de-noising digital images corrupted by impulse noise. The proposed method is based on cellular automata (CA) and fuzzy cellular automata (FCA). In this method, a given image ...
Read More
In this paper, a novel method named RHDINTPM (Removing High Density Impulse Noise via a Novel Two Phase Method) is proposed for de-noising digital images corrupted by impulse noise. The proposed method is based on cellular automata (CA) and fuzzy cellular automata (FCA). In this method, a given image is mapped to a CA. That is, every pixel of the image is associated with a cell of CA. RHDINTPM is composed of a two-phase filter. The first phase of the proposed method is a two-step noise detector so in the first step the corrupted pixels are diagnosed by the intensity of the minimum value and average Moore neighborhood pixels for central Pixel. In the second step, in order to increase accuracy in improving noise detection, the uncorrupted pixels remained from the first step are investigated by cellular automata. In the second phase of the method, the defective pixels of two-dimensional fuzzy cellular automata are restored using the structure of the Moore neighborhood. The experimental analysis demonstrates that the proposed filter is robust enough to very high levels of noise as high as 90% and preserves the meaningful detail of the image. Also, the proposed approach outperforms other representative filtering techniques in terms of image noise suppression and detail preservation.
Peyman Dehghanpour; Fardad Farokhi; Reza Rostami
Volume 07, Issue 02 , Spring 2018, , Pages 77-81
Abstract
Working memory (WM) is a part of human memory, the ability to maintain and manipulate information. WM performance is impaired in some neurological and psychiatric disorders such as schizophrenia and ADHD. Neurofeedback training is a self-regulation method which can be used to improve WM performance by ...
Read More
Working memory (WM) is a part of human memory, the ability to maintain and manipulate information. WM performance is impaired in some neurological and psychiatric disorders such as schizophrenia and ADHD. Neurofeedback training is a self-regulation method which can be used to improve WM performance by changing related EEG parameters. In this paper we used neurofeedback training to improve WM performance in eight healthy individuals. The protocol was consisted of individual upper alpha up-training in parietal brain lobe of participants which is a part of fronto-parietal network and related to central executive functions of WM. Power of individual upper alpha band in channels P3 and P4 was used for neurofeedback training in five sessions. 2-back working memory test was used to measure WM performance before and after the course. Results indicated success of subjects in neurofeedback training and enhancement of individual upper alpha power in both channels (P3 and P4). Results of 2-back test indicated that improvement in response accuracy and response time of test was significant. Also the correlation between the change in power of individual upper alpha band in channel P3 and change in response time of 2-back test was significant approximately (r= -0.571 and P=0.076). In conclusion it seems that individual upper alpha neurofeedback up-training in parietal lobe is an appropriate method to improve WM performance.
Armaghan Rahimi; fardad farokhi; Shahram Javadi
Volume 05, Issue 02 , Spring 2016, , Pages 93-99
Abstract
Smart home is composed of several controllers with different plants in control. If each controller works independently, without considering the mutual effect of the others in the control process, the whole system could definitely not converge to an optimum desired status and may not ever reach the demanded ...
Read More
Smart home is composed of several controllers with different plants in control. If each controller works independently, without considering the mutual effect of the others in the control process, the whole system could definitely not converge to an optimum desired status and may not ever reach the demanded condition. The function of different controller system may has conflict In some condition or emergency cases so the whole system could not converge to the steady state and receive to the demanded condition .According to the mentioned problem a new approach is presented in this paper using an integrated fuzzy controller in order to control light, temperature and emergency conditions in a realistic Matlab Simulink simulation. All environmental parameters that have effect on controlling different parts of smart home consider as an input parameter of system, such as Light, Heat and Smoke. The obtained results represent the presented approach and fuzzy model to be able and easy to implement the controller for smart home applications.
Iman Khandani; Reza Sabbaghi-Nadooshan; Fardad Farokhi
Volume 05, Issue 01 , Winter 2016, , Pages 49-52
Abstract
Wireless sensor networks(WSN) have unique properties that distinguished them from other wireless networks and have special challenges. Not-chargeable, not-changeable and limited power supplies of sensor nodes is the most important challenge of this networks, and if the power supply of node expired, a ...
Read More
Wireless sensor networks(WSN) have unique properties that distinguished them from other wireless networks and have special challenges. Not-chargeable, not-changeable and limited power supplies of sensor nodes is the most important challenge of this networks, and if the power supply of node expired, a part of data maybe lost. Because of the importance of covers in wireless sensors, in this work we have presented approaches that in their designs nodes are in special and different areas so that the hole of environment be covered and for saving power consumption, in each active period only one node is active and others are in inactive. For implementing the task cycle and avoiding from over-load of one node, the active node which have the task of sensing and maintaining the cover of area, in the next period, another node will replace it. It seems that by zoning the environment and cycle of tasks, the overall cover of wireless network improves.
Fardad Farrokhi; Mehdi Heydarian; Kaveh Kangarloo
Volume 03, Issue 02 , Spring 2014, , Pages 61-72
Abstract
This paper presents a comparison study between the multilayer perceptron (MLP) and radial basis function (RBF) neural networks with supervised learning and back propagation algorithm to track hand gestures. Both networks have two output classes which are hand and face. Skin is detected by a regional ...
Read More
This paper presents a comparison study between the multilayer perceptron (MLP) and radial basis function (RBF) neural networks with supervised learning and back propagation algorithm to track hand gestures. Both networks have two output classes which are hand and face. Skin is detected by a regional based algorithm in the image, and then networks are applied on video sequences frame by frame in different background (simple and complex) with different illumination of environment to detect face, hand and its gesture. The number of training and testing samples in networks are equal and the set of binary images obtained from skin detection method is used to train the networks. Hand gestures are 6 cases which are tracked and they were not recognized. Both left and right hands has been trained to the network. Network features are based on the image transforms and they should not relate to deformation, size and rotation of hand. Since some of the features are in common with each other so a new method is applied to reduced calculation of input vector. Result shows that MLP has high accuracy and higher speed in tracking hand gesture in different background with minimum average error but it has a lower speed in training and convergence compare to the RBF in its final average error.
Hamid Malmir; Fardad Farokhi; Reza Sabbaghi-Nadooshan
Volume 03, Issue 01 , Winter 2014, , Pages 47-53
Abstract
With the rapid development of the internet, the amount of information and data which are produced, are extremely massive. Hence, client will be confused with huge amount of data, and it is difficult to understand which ones are useful. Data mining can overcome this problem. While data mining is using ...
Read More
With the rapid development of the internet, the amount of information and data which are produced, are extremely massive. Hence, client will be confused with huge amount of data, and it is difficult to understand which ones are useful. Data mining can overcome this problem. While data mining is using on cloud computing, it is reducing time of processing, energy usage and costs. As the speed of data mining is very important, this paper proposes four faster classification algorithms in comparison with each other. In this paper, A Multi-Layer perceptron (MLP) Network is trained with Imperialist Competitive Algorithm (ICA), Particle Swarm Optimization (PSO), Differential Evolution (DE), and Invasive Weed Optimization (IWO) separately. The classifications are done on Wisconsin Breast Cancer (WBC) data base. At the end, to illustrate the speed and accuracy of these classifiers, they are compared with each other and two other types of Genetic algorithm classifiers (GA).
Roya Abdollahzadeh Badelbo; Fardad Farokhi; Alireza Kashaniniya
Volume 02, Issue 4 , Autumn 2013, , Pages 217-222
Abstract
In this article different types of artificial neural networks (ANN) were used for CNTFET (carbon nanotube transistors) simulation. CNTFET is one of the most likely alternatives to silicon transistors due to its excellent electronic properties. In determining the accurate output drain current of CNTFET, ...
Read More
In this article different types of artificial neural networks (ANN) were used for CNTFET (carbon nanotube transistors) simulation. CNTFET is one of the most likely alternatives to silicon transistors due to its excellent electronic properties. In determining the accurate output drain current of CNTFET, time lapsed and accuracy of different simulation methods were compared. The training data for ANNs were obtained by numerical ballistic FETToy model which is not directly applicable in circuit simulators like HSPICE. The ANN models were simulated in MATLAB R2010a software. In order to achieve more effective and consistent features, the UTA method was used and the overall performance of the models was tested in MATLAB. Finally the fast and accurate structure was introduced as a sub circuit for implementation in HSPICE simulator and then the implemented model was used to simulate a current source and an inverter circuit. Results indicate that the proposed ANN model is suitable for nanoscale circuits to be used in simulators like HSPICE.
Elham Yazdankhah; Fardad Farokhi; Reza Sabbaghi-Nadooshan
Volume 02, Issue 3 , Summer 2013, , Pages 173-180
Abstract
Wireless sensor networks, due to the characteristics of sensors such as wireless communication channels, the lack of infrastructure and targeted threats, are very vulnerable to the various attacks. Routing attacks on the networks, where a malicious node from sending data to the base station is perceived. ...
Read More
Wireless sensor networks, due to the characteristics of sensors such as wireless communication channels, the lack of infrastructure and targeted threats, are very vulnerable to the various attacks. Routing attacks on the networks, where a malicious node from sending data to the base station is perceived. In this article, a method that can be used to transfer the data securely to prevent attacks is suggested. The selection based on optimal path by routing using genetic algorithm uses. The proposed optimal paths to transmit data perceived to have chosen and ensures reliable data transmission.
Mahan Sedehzadeh; Farokhi Fardad
Volume 02, Issue 2 , Spring 2013, , Pages 77-85
Abstract
In this paper, a method for automatic stitching of radiology images based on pixel features has been presented. In this method, according to the smooth texture of radiological images and in order to increase the number of the extracted features after quality enhancement of initial radiology images, 45 ...
Read More
In this paper, a method for automatic stitching of radiology images based on pixel features has been presented. In this method, according to the smooth texture of radiological images and in order to increase the number of the extracted features after quality enhancement of initial radiology images, 45 degree isotropic mask is applied to each radiology image to observe the image details. After this process, we used statistical and heuristic image noise extraction method (SHINE) to acceptably reduce the noise resulting from radiation of alternating X-rays on detector. Pixel point’s features are obtained by selecting maximum or minimum value of the brightness of pixels in certain neighborhood of the resulting radiology images. This algorithm transmutes point’s features to 128 dimensional vector features. In order to identify the segments overlapping in basic radiology images, we specify equivalent vector features of each radiology image using the mathematical properties of the vectors and find the fit geometry transform between pairs features matched by the random sample consensus (RANSAC) algorithm. Finally, resulted motion model is applied to the initial radiology images and we stitch them together in a common surface
Saber Ghadri; Shahram Javadi; Fardad Farokhi
Volume 02, Issue 2 , Spring 2013, , Pages 91-94
Abstract
Today, DC motors is still being used globally due to their easy speed controllability. In this article, an Adaptive Neuro-Fuzzy Inference System (ANFIS) controller is designed for DC motors. The main purpose of performing such task is to reduce the DC motor starting current and deleting the ripple current ...
Read More
Today, DC motors is still being used globally due to their easy speed controllability. In this article, an Adaptive Neuro-Fuzzy Inference System (ANFIS) controller is designed for DC motors. The main purpose of performing such task is to reduce the DC motor starting current and deleting the ripple current during starting time in considering control parameters such as: rise time, settling time, maximum overshoot and system steady state error. The results have been simulated in MATLAB and a comparison is made between ANFIS controller and PID controller
Farnaz Rouhbakhsh; Fardad Farokhi; Kaveh Kangarloo
Volume 01, Issue 03 , Summer 2012, , Pages 199-204
Abstract
Since most common form of cervical cancer starts with pre-cancerous changes, a flawless detection of these changes becomes an important issue to prevent and treat the cervix cancer. There are 2 ways to stop this disease from developing. One way is to find and treat pre-cancers before they become true ...
Read More
Since most common form of cervical cancer starts with pre-cancerous changes, a flawless detection of these changes becomes an important issue to prevent and treat the cervix cancer. There are 2 ways to stop this disease from developing. One way is to find and treat pre-cancers before they become true cancers, and the other is to prevent the pre-cancers in the first place. The presented approach uses precancerous images which are taken from a digital colposcope, and a set of texture and color features is extracted which includes low and high grade SIL (Squamous Interepithelial Lesion ) .After extracting, features are fed to a classifier, which could be KNN,RBF,MLP and Neuro-Fuzzy network and after training effective features are selected using UTA algorithm for each classifier individually. Finally, results come in a comparison table, show that the landa fourteenth, theta-x and together with Neuro-fuzzy classifier have the best overall performance. This approach has an acceptable and simple early diagnosis of cervix cancer and may have found clinical application
