Yasar Becerikli | Kocaeli University (original) (raw)
Papers by Yasar Becerikli
Applied Mechanics and Materials, 2016
In the areas of surveillance and mobile communications, computer-aided evaluation of coverage are... more In the areas of surveillance and mobile communications, computer-aided evaluation of coverage area and measure of the field of sensors positioned on a created digital elevation model (DEM) is a facilitating method for positioning sensors and communication equipment. Fast evaluation of sensor coverage enables a computer to discover appropriate sensor locations by testing miscellaneous locations. In this paper, we describe a coverage maximization method that uses genetic algorithm and drawing algorithms based on integer arithmetic for coverage evaluation. Furthermore, solution of a sample problem and performance data gathered during solving process are presented, and the method's ability of finding good positions of radars for sea surveillance on a single processor PC in a duration less than ten minutes demonstrated.
International Conference of Computational Methods in Sciences and Engineering 2004 (ICCMSE 2004), 2019
2022 7th International Conference on Computer Science and Engineering (UBMK)
In this study, linear programming and integer programming methods and their application areas are... more In this study, linear programming and integer programming methods and their application areas are discussed, and these methods are implemented to stock cutting problem. In the application a new approach is used on solution of two dimensional stock cutting problem and performance of new method is compared to the other methods. The aim of application is to obtain minimal leakage with fastest solution by using sizes, quantities of resource and target stocks, plotting the layout on screen.
Turkiye Bilişim Vakfi Bilgisayar Bilimleri Ve Muhendisliği Dergisi, 2012
Levenberg-Marquardt (LM) algoritmasi yapay sinir aglarinin egitiminde saglamis oldugu hiz ve kara... more Levenberg-Marquardt (LM) algoritmasi yapay sinir aglarinin egitiminde saglamis oldugu hiz ve kararlilik nedeni ile tercih edilmektedir. Bu calismada yapay sinir agi (YSA) egitiminin LM algoritmasi ile kayan noktali sayi formatinda donanimsal olarak FPGA’da gerceklenmesi sunulmustur. Donanimsal gercekleme ISE Webpack10.1 programi kullanilarak Xilinx Virtex 5 xc5vlx110-3ff1153 FPGA’si uzerinde gerceklenmistir. Calismada ozellikle YSA mimarisinin dogasinda var olan paralelligin FPGA uzerine aktarilmasinin yani sira egitim asamasinda LM algoritmasi da paralel veri islemeye uygun olarak gerceklenmistir. Elde edilen sentez sonuclari, LM ile YSA egitiminin FPGA uzerinde basari ile gerceklenebilecegini ortaya koymustur Abstract Hardware Implementation of Neural Network Training with Levenberg-Marquardt Algorithm Levenberg-Marquardt (LM) algorithm is preferred due to providing fast convergence and stability in training of artificial neural networks (ANN). In this study, hardware implementation of ANN training with LM algorithm is presented on FPGA using floating point number representation. . Hardware implementation has been realized on Virtex-5 xc5vlx110-3ff1153 FPGA using ISE Webpack 10.1 software. In this work, both ANN and its training using LM have been particularly implemented on FPGA according to the inherent parallel data processing of ANN. Obtained synthesis results have showed that training of ANN using LM algorithm can be successfully implemented on FPGA.
Neural Networks, 2003
The application of neural networks technology to dynamic system control has been constrained by t... more The application of neural networks technology to dynamic system control has been constrained by the non-dynamic nature of popular network architectures. Many of difficulties are-large network sizes (i.e. curse of dimensionality), long training times, etc. These problems can be overcome with dynamic neural networks (DNN). In this study, intelligent optimal control problem is considered as a nonlinear optimization with dynamic equality constraints, and DNN as a control trajectory priming system. The resulting algorithm operates as an auto-trainer for DNN (a self-learning structure) and generates optimal feed-forward control trajectories in a significantly smaller number of iterations. In this way, optimal control trajectories are encapsulated and generalized by DNN. The time varying optimal feedback gains are also generated along the trajectory as byproducts. Speeding up trajectory calculations opens up avenues for real-time intelligent optimal control with virtual global feedback. We used direct-descent-curvature algorithm with some modifications (we called modified-descend-controller-MDC algorithm) for the optimal control computations. The algorithm has generated numerically very robust solutions with respect to conjugate points. The adjoint theory has been used in the training of DNN which is considered as a quasi-linear dynamic system. The updating of weights (identification of parameters) are based on Broyden-Fletcher-Goldfarb-Shanno BFGS method. Simulation results are given for an intelligent optimal control system controlling a difficult nonlinear second-order system using fully connected three-neuron DNN.
Mathematical and Computer Modelling, 2007
In this paper, a fuzzy controller for an inverted pendulum system is presented in two stages. The... more In this paper, a fuzzy controller for an inverted pendulum system is presented in two stages. These stages are: investigation of fuzzy control system modeling methods and solution of the "Inverted Pendulum Problem" by using Java programming with Applets for internet based control education. In the first stage, fuzzy modeling and fuzzy control system investigation, Java programming language, classes and multithreading were introduced. In the second stage specifically, simulation of the inverted pendulum problem was developed with Java Applets and the simulation results were given. Also some stability concepts are introduced.
ISA Transactions, 2006
A nonlinear predictive control technique is developed to determine the optimal drying profile for... more A nonlinear predictive control technique is developed to determine the optimal drying profile for a drying process. A complete nonlinear model of the baker's yeast drying process is used for predicting the future control actions. To minimize the difference between the model predictions and the desired trajectory throughout finite horizan, an objective function is described. The optimization problem is solved using a genetic algorithm due to the successful overconventional optimization techniques in the applications of the complex optimization problems. The control scheme comprises a drying process, a nonlinear prediction model, an optimizer, and a genetic search block. The nonlinear predictive control method proposed in this paper is applied to the baker's yeast drying process. The results show significant enhancement of the manufacturing quality, considerable decrease of the energy consumption and drying time, obtained by the proposed nonlinear predictive control.
ISA Transactions, 2006
This paper presents a direct descent second order or direct descent curvature algorithm with some... more This paper presents a direct descent second order or direct descent curvature algorithm with some modifications for the optimal control computations. This algorithm is compared with Hamiltonian methods in the literature. The proposed algorithm has generated numerically robust solutions with respect to conjugate points. The weighting matrix updating scheme was developed to improve the second-order optimal control algorithm, tested the performance of the algorithm, and shown on the benchmark and industrial process. The time-varying optimal feedback (TVOFB) gains are also generated along the trajectory as byproducts. If the trajectory deviates from the optimal trajectory for any reason (i.e., changing of system parameters, step disturbance into the plant, changing of initial conditions), it is held on the optimal trajectory by means of the optimal feedback. Simulations have been given for controlling the Van der Pol and bioreactor system, which are nonlinear benchmark systems.
The Canadian Journal of Chemical Engineering, 2008
A distributed parameter model was developed to predict the drying behaviour of granular baker's y... more A distributed parameter model was developed to predict the drying behaviour of granular baker's yeast by setting up material and heat balances at the particle level. Temperature and moisture gradients were calculated for cylindrical and spherical granules. The performance of the model with two granule sizes was compared with experimental measurements. The model was initially used for non-shrinking granules but later modified to take shrinkage into account. The reduction in granule size during the course of drying was estimated and good correspondence with experimental measurements was obtained. In addition to temperature and moisture gradients, the product quality was predicted during drying and compared to experimental results. The accuracy of the model was better than the lumped parameter model. On a mis au point un modèleà paramètres distribués afin de prédire le séchage de la levure de boulanger granulaire enétablissant les bilans de matière et de chaleur au niveau des particules. Les gradients de température et d'humidité ontété calculés pour des granules cylindriques et sphériques. La performance du modèle avec les deux tailles de granules est comparée aux mesures expérimentales. Le modèle a d'abordété utilisé pour les granules indéformables mais aété ensuite adapté pour prendre le rétrécissement en compte. On a estimé la réduction de taille des granules lors du séchage et un bon accord aété trouvé avec les mesures expérimentales. Outre les gradients de température et d'humidité, la qualité de produit aété prédite lors du séchage et comparée aux résultats expérimentaux. La précision du modèle est meilleure que pour le modèleà paramètres regroupés.
ktokai-u.ac.jp
The edge detection is one of the most important tasks in the image processing area. The edges in ... more The edge detection is one of the most important tasks in the image processing area. The edges in the image can be defined as sudden gray level transition. In other words edges are the high frequency components of the images. At the same time, random noises have also same attributes. Thus distinguishing the edges from the noises is one of the most complex problems that is encountered in the image processing. In the literature some techniques and approaches are available for this purpose. In this paper we proposed a new fuzzy based approach to fulfill filtering and edge extraction simultaneously. It was proposed a new approach for filtering and edge detection using fuzzy approach. In these approaches, heuristic rules were applied, and results were observed for different images.
IEEE Access, 2020
Changes and progresses in information technologies have played an important role in the developme... more Changes and progresses in information technologies have played an important role in the development of intelligent vehicle systems in recent years. Driver fatigue is an important factor in vehicle accidents. For this reason, traffic accidents involving driver fatigue and driver carelessness have been followed by researchers. In this article, a Multi-tasking Convulational Neural Network (ConNN *) model is proposed to detect driver drowsiness/fatigue. Eye and mouth characteristics are utilized for driver's behavior model. Changes to these characteristics are used to monitor driver fatigue. With the proposed Multi-task ConNN model, unlike the studies in the literature, both mouth and eye information are classified into a single model at the same time. Driver fatigue is determined by calculating eyes' closure duration/Percentage of eye closure (PERCLOS) and yawning frequency/frequency of mouth (FOM). In this study, the fatigue degree of the driver is divided into 3 classes. The proposed model achieved 98.81% fatigue detection on YawdDD and NthuDDD dataset. The success of the model is presented comparatively.
Intelligent Automation & Soft Computing, 2022
Convolutional Neural Network (ConNN) implementations on Field Programmable Gate Array (FPGA) are ... more Convolutional Neural Network (ConNN) implementations on Field Programmable Gate Array (FPGA) are being studied since the computational capabilities of FPGA have been improved recently. Model compression is required to enable ConNN deployment on resource-constrained FPGA devices. Logarithmic quantization is one of the efficient compression methods that can compress a model to very low bit-width without significant deterioration in performance. It is also hardware-friendly by using bitwise operations for multiplication. However, the logarithmic suffers from low resolution at high inputs due to exponential properties. Therefore, we propose a modified logarithmic quantization method with a fine resolution to compress a neural network model. In experiments, quantized models achieve a negligible loss of accuracy without the need for retraining steps. Besides this, we propose a resource-efficient hardware accelerator for running ConNN inference. Our design completely eliminates multipliers with bit shifters and adders. Throughput is measured in Giga Operation Per Second (GOP/s). The hardware utilization efficiency is represented by GOP/s per block of Digital Signal Processing (DSP) and Look-up Tables (LUTs). The result shows that the accelerator achieves resource efficiency of 9.38 GOP/s/DSP and 3.33 GOP/s/kLUTs.
2017 10th International Conference on Electrical and Electronics Engineering (ELECO), 2017
Human or pedestrian detection is an attractive headline and has been proposed in computer vision ... more Human or pedestrian detection is an attractive headline and has been proposed in computer vision and machine learning fields. Real time detection and low power system is a critical challenges. Support Vector Machine algorithm with Histograms of oriented gradients (HOG) feature descriptor is given a high successful result, fast and reliable, for human detection. Therefore, this paper demonstrates how to implement HOG feature descriptor with Support Vector Machine (SVM) using FPGA and presents a report that includes FPGA's resource utilization, time consuming, power consumption and SVM accuracy results.
Expert Systems, 2010
: Estimation of the plaque area in intravascular ultrasound images after extraction of the media ... more : Estimation of the plaque area in intravascular ultrasound images after extraction of the media and plaque–lumen interfaces is an important application of computer-aided diagnosis in medical imaging. This paper presents a novel system for fully automatic and fast calculation of plaque quantity by capturing the surrounding ring called media. The system utilizes an algorithm that consists of an enhanced technique for noise removal and a method of detecting different iso levels by sinking the image gradually under zero level. Moreover, an important novelty with this technique is the simultaneous extraction of media and lumen–plaque interfaces at satisfactory levels. There are no higher dimensional surfaces and evolution of contours, stopping at high image gradients. Thus, the system runs really fast with curvature velocity only and has no complexity. Experiments also show that this shape-recovering curvature term not only removes the noisy behaviour of ultrasound images but also strengthens very weak boundaries and even completes the missing walls of the media. In addition, the lumen–plaque interface can be detected simultaneously. For validation, a new and very useful algorithm is developed for labelling of intravascular ultrasound images, taken from video sequences of 15 patients, and a comparison-based verification is done between manual contours by experts and the contours extracted by our system.
2011 XXIII International Symposium on Information, Communication and Automation Technologies, 2011
Automatic quantitative analysis of brain tissues has a high importance. However, lack of high pre... more Automatic quantitative analysis of brain tissues has a high importance. However, lack of high precision still makes results unreliable. This paper presents a novel system which separates cortical GM from WM. It imitates human perception like edge detection algorithms but recovers their disability in segmentation. System is fully automatic and unsupervised. Fastened segments of geodesic passive contours (FSG) are utilized and the perceptive sensitivity to edges is imitated. This nature of the solution proved to treat the inhomogeneity and noise problems well. The technique is tested on both real and synthetic databases and compared with widely used software of SPM and works faster. Our technique succeeded in getting average misclassification rate of 4.8% for WM and correct GM-WM boundary rate of 77% being very close to experts' agreement.
Applied Mechanics and Materials, 2016
In the areas of surveillance and mobile communications, computer-aided evaluation of coverage are... more In the areas of surveillance and mobile communications, computer-aided evaluation of coverage area and measure of the field of sensors positioned on a created digital elevation model (DEM) is a facilitating method for positioning sensors and communication equipment. Fast evaluation of sensor coverage enables a computer to discover appropriate sensor locations by testing miscellaneous locations. In this paper, we describe a coverage maximization method that uses genetic algorithm and drawing algorithms based on integer arithmetic for coverage evaluation. Furthermore, solution of a sample problem and performance data gathered during solving process are presented, and the method's ability of finding good positions of radars for sea surveillance on a single processor PC in a duration less than ten minutes demonstrated.
International Conference of Computational Methods in Sciences and Engineering 2004 (ICCMSE 2004), 2019
2022 7th International Conference on Computer Science and Engineering (UBMK)
In this study, linear programming and integer programming methods and their application areas are... more In this study, linear programming and integer programming methods and their application areas are discussed, and these methods are implemented to stock cutting problem. In the application a new approach is used on solution of two dimensional stock cutting problem and performance of new method is compared to the other methods. The aim of application is to obtain minimal leakage with fastest solution by using sizes, quantities of resource and target stocks, plotting the layout on screen.
Turkiye Bilişim Vakfi Bilgisayar Bilimleri Ve Muhendisliği Dergisi, 2012
Levenberg-Marquardt (LM) algoritmasi yapay sinir aglarinin egitiminde saglamis oldugu hiz ve kara... more Levenberg-Marquardt (LM) algoritmasi yapay sinir aglarinin egitiminde saglamis oldugu hiz ve kararlilik nedeni ile tercih edilmektedir. Bu calismada yapay sinir agi (YSA) egitiminin LM algoritmasi ile kayan noktali sayi formatinda donanimsal olarak FPGA’da gerceklenmesi sunulmustur. Donanimsal gercekleme ISE Webpack10.1 programi kullanilarak Xilinx Virtex 5 xc5vlx110-3ff1153 FPGA’si uzerinde gerceklenmistir. Calismada ozellikle YSA mimarisinin dogasinda var olan paralelligin FPGA uzerine aktarilmasinin yani sira egitim asamasinda LM algoritmasi da paralel veri islemeye uygun olarak gerceklenmistir. Elde edilen sentez sonuclari, LM ile YSA egitiminin FPGA uzerinde basari ile gerceklenebilecegini ortaya koymustur Abstract Hardware Implementation of Neural Network Training with Levenberg-Marquardt Algorithm Levenberg-Marquardt (LM) algorithm is preferred due to providing fast convergence and stability in training of artificial neural networks (ANN). In this study, hardware implementation of ANN training with LM algorithm is presented on FPGA using floating point number representation. . Hardware implementation has been realized on Virtex-5 xc5vlx110-3ff1153 FPGA using ISE Webpack 10.1 software. In this work, both ANN and its training using LM have been particularly implemented on FPGA according to the inherent parallel data processing of ANN. Obtained synthesis results have showed that training of ANN using LM algorithm can be successfully implemented on FPGA.
Neural Networks, 2003
The application of neural networks technology to dynamic system control has been constrained by t... more The application of neural networks technology to dynamic system control has been constrained by the non-dynamic nature of popular network architectures. Many of difficulties are-large network sizes (i.e. curse of dimensionality), long training times, etc. These problems can be overcome with dynamic neural networks (DNN). In this study, intelligent optimal control problem is considered as a nonlinear optimization with dynamic equality constraints, and DNN as a control trajectory priming system. The resulting algorithm operates as an auto-trainer for DNN (a self-learning structure) and generates optimal feed-forward control trajectories in a significantly smaller number of iterations. In this way, optimal control trajectories are encapsulated and generalized by DNN. The time varying optimal feedback gains are also generated along the trajectory as byproducts. Speeding up trajectory calculations opens up avenues for real-time intelligent optimal control with virtual global feedback. We used direct-descent-curvature algorithm with some modifications (we called modified-descend-controller-MDC algorithm) for the optimal control computations. The algorithm has generated numerically very robust solutions with respect to conjugate points. The adjoint theory has been used in the training of DNN which is considered as a quasi-linear dynamic system. The updating of weights (identification of parameters) are based on Broyden-Fletcher-Goldfarb-Shanno BFGS method. Simulation results are given for an intelligent optimal control system controlling a difficult nonlinear second-order system using fully connected three-neuron DNN.
Mathematical and Computer Modelling, 2007
In this paper, a fuzzy controller for an inverted pendulum system is presented in two stages. The... more In this paper, a fuzzy controller for an inverted pendulum system is presented in two stages. These stages are: investigation of fuzzy control system modeling methods and solution of the "Inverted Pendulum Problem" by using Java programming with Applets for internet based control education. In the first stage, fuzzy modeling and fuzzy control system investigation, Java programming language, classes and multithreading were introduced. In the second stage specifically, simulation of the inverted pendulum problem was developed with Java Applets and the simulation results were given. Also some stability concepts are introduced.
ISA Transactions, 2006
A nonlinear predictive control technique is developed to determine the optimal drying profile for... more A nonlinear predictive control technique is developed to determine the optimal drying profile for a drying process. A complete nonlinear model of the baker's yeast drying process is used for predicting the future control actions. To minimize the difference between the model predictions and the desired trajectory throughout finite horizan, an objective function is described. The optimization problem is solved using a genetic algorithm due to the successful overconventional optimization techniques in the applications of the complex optimization problems. The control scheme comprises a drying process, a nonlinear prediction model, an optimizer, and a genetic search block. The nonlinear predictive control method proposed in this paper is applied to the baker's yeast drying process. The results show significant enhancement of the manufacturing quality, considerable decrease of the energy consumption and drying time, obtained by the proposed nonlinear predictive control.
ISA Transactions, 2006
This paper presents a direct descent second order or direct descent curvature algorithm with some... more This paper presents a direct descent second order or direct descent curvature algorithm with some modifications for the optimal control computations. This algorithm is compared with Hamiltonian methods in the literature. The proposed algorithm has generated numerically robust solutions with respect to conjugate points. The weighting matrix updating scheme was developed to improve the second-order optimal control algorithm, tested the performance of the algorithm, and shown on the benchmark and industrial process. The time-varying optimal feedback (TVOFB) gains are also generated along the trajectory as byproducts. If the trajectory deviates from the optimal trajectory for any reason (i.e., changing of system parameters, step disturbance into the plant, changing of initial conditions), it is held on the optimal trajectory by means of the optimal feedback. Simulations have been given for controlling the Van der Pol and bioreactor system, which are nonlinear benchmark systems.
The Canadian Journal of Chemical Engineering, 2008
A distributed parameter model was developed to predict the drying behaviour of granular baker's y... more A distributed parameter model was developed to predict the drying behaviour of granular baker's yeast by setting up material and heat balances at the particle level. Temperature and moisture gradients were calculated for cylindrical and spherical granules. The performance of the model with two granule sizes was compared with experimental measurements. The model was initially used for non-shrinking granules but later modified to take shrinkage into account. The reduction in granule size during the course of drying was estimated and good correspondence with experimental measurements was obtained. In addition to temperature and moisture gradients, the product quality was predicted during drying and compared to experimental results. The accuracy of the model was better than the lumped parameter model. On a mis au point un modèleà paramètres distribués afin de prédire le séchage de la levure de boulanger granulaire enétablissant les bilans de matière et de chaleur au niveau des particules. Les gradients de température et d'humidité ontété calculés pour des granules cylindriques et sphériques. La performance du modèle avec les deux tailles de granules est comparée aux mesures expérimentales. Le modèle a d'abordété utilisé pour les granules indéformables mais aété ensuite adapté pour prendre le rétrécissement en compte. On a estimé la réduction de taille des granules lors du séchage et un bon accord aété trouvé avec les mesures expérimentales. Outre les gradients de température et d'humidité, la qualité de produit aété prédite lors du séchage et comparée aux résultats expérimentaux. La précision du modèle est meilleure que pour le modèleà paramètres regroupés.
ktokai-u.ac.jp
The edge detection is one of the most important tasks in the image processing area. The edges in ... more The edge detection is one of the most important tasks in the image processing area. The edges in the image can be defined as sudden gray level transition. In other words edges are the high frequency components of the images. At the same time, random noises have also same attributes. Thus distinguishing the edges from the noises is one of the most complex problems that is encountered in the image processing. In the literature some techniques and approaches are available for this purpose. In this paper we proposed a new fuzzy based approach to fulfill filtering and edge extraction simultaneously. It was proposed a new approach for filtering and edge detection using fuzzy approach. In these approaches, heuristic rules were applied, and results were observed for different images.
IEEE Access, 2020
Changes and progresses in information technologies have played an important role in the developme... more Changes and progresses in information technologies have played an important role in the development of intelligent vehicle systems in recent years. Driver fatigue is an important factor in vehicle accidents. For this reason, traffic accidents involving driver fatigue and driver carelessness have been followed by researchers. In this article, a Multi-tasking Convulational Neural Network (ConNN *) model is proposed to detect driver drowsiness/fatigue. Eye and mouth characteristics are utilized for driver's behavior model. Changes to these characteristics are used to monitor driver fatigue. With the proposed Multi-task ConNN model, unlike the studies in the literature, both mouth and eye information are classified into a single model at the same time. Driver fatigue is determined by calculating eyes' closure duration/Percentage of eye closure (PERCLOS) and yawning frequency/frequency of mouth (FOM). In this study, the fatigue degree of the driver is divided into 3 classes. The proposed model achieved 98.81% fatigue detection on YawdDD and NthuDDD dataset. The success of the model is presented comparatively.
Intelligent Automation & Soft Computing, 2022
Convolutional Neural Network (ConNN) implementations on Field Programmable Gate Array (FPGA) are ... more Convolutional Neural Network (ConNN) implementations on Field Programmable Gate Array (FPGA) are being studied since the computational capabilities of FPGA have been improved recently. Model compression is required to enable ConNN deployment on resource-constrained FPGA devices. Logarithmic quantization is one of the efficient compression methods that can compress a model to very low bit-width without significant deterioration in performance. It is also hardware-friendly by using bitwise operations for multiplication. However, the logarithmic suffers from low resolution at high inputs due to exponential properties. Therefore, we propose a modified logarithmic quantization method with a fine resolution to compress a neural network model. In experiments, quantized models achieve a negligible loss of accuracy without the need for retraining steps. Besides this, we propose a resource-efficient hardware accelerator for running ConNN inference. Our design completely eliminates multipliers with bit shifters and adders. Throughput is measured in Giga Operation Per Second (GOP/s). The hardware utilization efficiency is represented by GOP/s per block of Digital Signal Processing (DSP) and Look-up Tables (LUTs). The result shows that the accelerator achieves resource efficiency of 9.38 GOP/s/DSP and 3.33 GOP/s/kLUTs.
2017 10th International Conference on Electrical and Electronics Engineering (ELECO), 2017
Human or pedestrian detection is an attractive headline and has been proposed in computer vision ... more Human or pedestrian detection is an attractive headline and has been proposed in computer vision and machine learning fields. Real time detection and low power system is a critical challenges. Support Vector Machine algorithm with Histograms of oriented gradients (HOG) feature descriptor is given a high successful result, fast and reliable, for human detection. Therefore, this paper demonstrates how to implement HOG feature descriptor with Support Vector Machine (SVM) using FPGA and presents a report that includes FPGA's resource utilization, time consuming, power consumption and SVM accuracy results.
Expert Systems, 2010
: Estimation of the plaque area in intravascular ultrasound images after extraction of the media ... more : Estimation of the plaque area in intravascular ultrasound images after extraction of the media and plaque–lumen interfaces is an important application of computer-aided diagnosis in medical imaging. This paper presents a novel system for fully automatic and fast calculation of plaque quantity by capturing the surrounding ring called media. The system utilizes an algorithm that consists of an enhanced technique for noise removal and a method of detecting different iso levels by sinking the image gradually under zero level. Moreover, an important novelty with this technique is the simultaneous extraction of media and lumen–plaque interfaces at satisfactory levels. There are no higher dimensional surfaces and evolution of contours, stopping at high image gradients. Thus, the system runs really fast with curvature velocity only and has no complexity. Experiments also show that this shape-recovering curvature term not only removes the noisy behaviour of ultrasound images but also strengthens very weak boundaries and even completes the missing walls of the media. In addition, the lumen–plaque interface can be detected simultaneously. For validation, a new and very useful algorithm is developed for labelling of intravascular ultrasound images, taken from video sequences of 15 patients, and a comparison-based verification is done between manual contours by experts and the contours extracted by our system.
2011 XXIII International Symposium on Information, Communication and Automation Technologies, 2011
Automatic quantitative analysis of brain tissues has a high importance. However, lack of high pre... more Automatic quantitative analysis of brain tissues has a high importance. However, lack of high precision still makes results unreliable. This paper presents a novel system which separates cortical GM from WM. It imitates human perception like edge detection algorithms but recovers their disability in segmentation. System is fully automatic and unsupervised. Fastened segments of geodesic passive contours (FSG) are utilized and the perceptive sensitivity to edges is imitated. This nature of the solution proved to treat the inhomogeneity and noise problems well. The technique is tested on both real and synthetic databases and compared with widely used software of SPM and works faster. Our technique succeeded in getting average misclassification rate of 4.8% for WM and correct GM-WM boundary rate of 77% being very close to experts' agreement.