Yazar "Donmez, Emrah" seçeneğine göre listele
Listeleniyor 1 - 16 / 16
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe Assessment of environmental factors affecting software reliability: a survey study(Tubitak Scientific & Technological Research Council Turkey, 2020) Ozcan, Alper; Catal, Cagatay; Togay, Cengiz; Tekinerdogan, Bedir; Donmez, EmrahCurrently, many systems depend on software, and software reliability as such has become one of the key challenges. Several studies have been carried out that focus on the impact of external environmental factors that impact software reliability. These studies, however, were all carried out in the same geographical context. Given the rapid developments in software engineering, this study aims to identify and reinvestigate the environmental factors that impact software reliability by also considering a different context. The environmental factors that have an impact on software reliability as reported in earlier studies have been analyzed and synthesized. Subsequently, a survey study is conducted to analyze the impact of 32 environmental factors from the perspective of multiple stakeholders. Several statistical analysis methods were applied for the analysis. Data were collected from 24 organizations and 70 software professionals. Most factors shown in top 10 lists of previous studies remain in the top 10 in our study, but their order is different. Testing coverage is now the most significant factor and testing effort is considered as the second most significant factor. The environmental factors defined previously retain their impact. The ordering of the importance of the environmental factors has changed though.Öğe Bi-RRT Path Extraction and Curve Fitting Smooth with Visual Based Configuration Space Mapping(Ieee, 2017) Donmez, Emrah; Kocamaz, A. Fatih; Dirik, MahmutPath planning is the one of the most basic research areas in robotics. It simply concern about acquiring a safe path with admissible cost. In this study, we adapt bidirectional rapidly random exploring tree (Bi-RRT) path extraction to visual based configuration space map hosting obstacles and smooth result path with curve fitting models. Firstly, a map of the configuration space is created and robot, target positions are detected with threshold based object detection. There are two positions where two distinct RRT are launched on this map. These positions are robot initial position and target position. Both RRT try to reach target with random branches in each iterations. When one of these RRT branch intersect with other RRT branch, the algorithm is stopped. The acquired trajectory is the path between initial position and target position. But acquired path is generally close to the obstacles and unnecessary branches or jagged parts can be formed. Therefore, to provide safety object dilation over obstacles are used. Finally, the path is smoothed with curve fitting models. We conduct several experiments to evaluate Bi-RRT performance.Öğe A Cloudware Architecture for Collaboration of Multiple AGVs in Indoor Logistics: Case Study in Fabric Manufacturing Enterprises(Mdpi, 2020) Okumus, Fatih; Donmez, Emrah; Kocamaz, Adnan FatihIn Industry 4.0 compatible workshops, the demand for Automated Guided Vehicles (AGVs) used in indoor logistics systems has increased remarkably. In these indoor logistics systems, it may be necessary to execute multiple transport tasks simultaneously using multiple AGVs. However, some challenges require special solutions for AGVs to be used in industrial autonomous transportation. These challenges can be addressed under four main headings: positioning, optimum path planning, collision avoidance and optimum task allocation. The solutions produced for these challenges may require special studies that vary depending on the type of tasks and the working environment in which AGVs are used. This study focuses on the problem of automated indoor logistics carried out in the simultaneous production of textile finishing enterprises. In the study, a centralized cloud system that enables multiple AGVs to work in collaboration has been developed. The finishing enterprise of a denim manufacturing factory was handled as a case study and modelling of mapping-planning processes was carried out using the developed cloud system. In the cloud system, RestFul APIs, for mapping the environment, and WebSocket methods, to track the locations of AGVs, have been developed. A collaboration module in harmony with the working model has been developed for AGVs to be used for fabric transportation. The collaboration module consists of task definition, battery management-optimization, selection of the most suitable batch trolleys (provides mobility of fabrics for the finishing mills), optimum task distribution and collision avoidance stages. In the collaboration module, all the finishing processes until the product arrives the delivery point are defined as tasks. A task allocation algorithm has been developed for the optimum performance of these tasks. The multi-fitness function that optimizes the total path of the AGVs, the elapsed time and the energy spent while performing the tasks have been determined. An assignment matrix based on K nearest neighbor (k-NN) and permutation possibilities was created for the optimal task allocation, and the most appropriate row was selected according to the optimal path totals of each row in the matrix. The D* Lite algorithm has been used to calculate the optimum path between AGVs and goals by avoiding static obstacles. By developing simulation software, the problem model was adapted and the operation of the cloud system was tested. Simulation results showed that the developed cloud system was successfully implemented. Although the developed cloud system has been applied as a case study in fabric finishing workshops with a complex structure, it can be used in different sectors as its logistic processes are similar.Öğe Design of Mobile Robot Control Infrastructure Based on Decision Trees and Adaptive Potential Area Methods(Springer Int Publ Ag, 2020) Donmez, Emrah; Kocamaz, Adnan FatihThere have been a great number of studies in the scope of mobile robot systems. The most critical tasks in these systems are control and path planning. The main goal of the control task is to develop a stable control system. On the other hand, the basic motivation in the path planning task is to find a safe path with an acceptable cost. In most researches, a moving robot is considered as a point mass object and only the simulation experiments are applied. In this study, a decision tree-based mobile robot control has been developed for a static indoor environment hosting obstacle(s). The camera has been hung vertically (eye-out-device configuration) to obtain the configuration area map and track the wheeled mobile robot (WMR). A suitable path plan has been extracted with the adaptive artificial potential field (APF) method on the image obtained from the camera. Virtual distance sensors are used to calculate the potentials for APF. A decision tree-based controller has been developed to model the motion characteristics of the robot. A trigonometry-based approach is used to calculate the controller inputs. The controller has steered the WMR on the path in real time. Both simulation and real-world experiments have been conducted on a WMR in different configuration spaces. It has been determined that the designed system is convenient for controlling the WMR. The data obtained are compared to show the difference between the desired and actual path planning results. The efficiency of the controller method has been greatly improved by using dynamic parameters in the control modules.Öğe Discrimination of Haploid and Diploid Maize Seeds Based on Deep Features(Ieee, 2020) Donmez, EmrahOne of the main processes in in vivo maternal haploid breeding method which is widely used in hybrid maize breeding in recent years is the separation of haploid and diploid individuals. Although different approaches have been proposed to make this distinction, the R1-nj color marker is widely and successfully used. The R1-nj color marker allows the visual separation of haploid and diploid individuals during the seed period. Nowadays, this distinction is done manually, causing loss of time and labor as well as high error. In this study, an open access dataset consisting of 3,000 maize seed images was used. The deep features from the FC6, FC7 and FC8 fully connected layers of the AlexNet architecture are classified with the support vector machine. 10-fold cross-validation test was used to evaluate model performances. Experimental results showed that the best classification performance is possible with 89.50% accuracy using deep features obtained from the FC6 fully connected layer.Öğe The Eye-out-Device Multi-Camera Expansion for Mobile Robot Control(Ieee, 2019) Donmez, Emrah; Kocamaz, Adnan FatihIn visual based control (VBC), a single eye-out-device camera provides a limited viewing area. The multi-camera configuration can be used to overcome this problem. However, simultaneous multi-image processing is a challenging task. In this study; a mobile robot control is performed with Gaussian and adaptive potential field based methods under the eye-out device multi-camera configuration. Images taken from the cameras are stitched according to common features. The color based object detection is operated to detect the robot, target and obstacles in this image. To acquire a suitable path between robot and target, adaptive potential field algorithm is executed. The Gaussian based mobile robot controller is used to drive on the robot according to the path plan. In this way, configuration space can be expanded via cameras. At the same time, systematic and unsystematic errors are avoided through VBC. Simulation and real-world experiments show that the system demonstrates a good performance and efficiency.Öğe A Hog & Graph Based Human Segmentation from Video Sequences(Ieee, 2018) Donmez, Emrah; Kocamaz, Adnan FatihHuman segmentation from video is a significant problem to recognize a specific person which is desired to find. There are a remarkable number of studies discussing segmentation process in videos. Almost all study is examined how well a human segmentation process could be created by considering their position, clothing and face. Unfortunately, it has been assumed that the background is static by approximately all of these works. Main challenges in this research area are; non-static background derived from dynamic structure of videos, human body and face position, shadows and clothing changing after a period of time in videos. In this script, we have proposed a multi-model (Histogram of the Gradients - HOG and Graph-based) human segmentation technique that has worked with respect to HOG features and detected low-level and mid-level features of human clothing by supposing their positions in video. The offered technique is designed to demonstrate robustness against such challenges emphasized above. In this study, a well-known video series have been used. The video scenes can reach 15 similar to 25 fps and have the size about 640x480px. To compare graph-based method robustness a well-known segmentation method Watershed is also experimented and both methods are simply compared. Eventually, we determined that the proposed technique can produce satisfactory quality segmentation mentioned and detailed in following sections.Öğe A hybrid DNN-LSTM model for detecting phishing URLs(Springer London Ltd, 2023) Ozcan, Alper; Catal, Cagatay; Donmez, Emrah; Senturk, BehcetPhishing is an attack targeting to imitate the official websites of corporations such as banks, e-commerce, financial institutions, and governmental institutions. Phishing websites aim to access and retrieve users' important information such as personal identification, social security number, password, e-mail, credit card, and other account information. Several anti-phishing techniques have been developed to cope with the increasing number of phishing attacks so far. Machine learning and particularly, deep learning algorithms are nowadays the most crucial techniques used to detect and prevent phishing attacks because of their strong learning abilities on massive datasets and their state-of-the-art results in many classification problems. Previously, two types of feature extraction techniques [i.e., character embedding-based and manual natural language processing (NLP) feature extraction] were used in isolation. However, researchers did not consolidate these features and therefore, the performance was not remarkable. Unlike previous works, our study presented an approach that utilizes both feature extraction techniques. We discussed how to combine these feature extraction techniques to fully utilize from the available data. This paper proposes hybrid deep learning models based on long short-term memory and deep neural network algorithms for detecting phishing uniform resource locator and evaluates the performance of the models on phishing datasets. The proposed hybrid deep learning models utilize both character embedding and NLP features, thereby simultaneously exploiting deep connections between characters and revealing NLP-based high-level connections. Experimental results showed that the proposed models achieve superior performance than the other phishing detection models in terms of accuracy metric.Öğe Multi Target Task Distribution and Path Planning for Multi-Agents(Ieee, 2018) Donmez, Emrah; Kocamaz, Adnan FatihIn the field of robotics, there are basic subjects such as control design, machine vision, path planning, performing assigned tasks. It is widely focused on one robot systems in literature. There are also studies on multiple robots and multiple target / task sharing. In this study, a task sharing system for navigating multiple targets with multiple robots and a navigation algorithm for finding the appropriate route has been investigated. This study is similar with respect to the problem of MultiTraveling Salesman Problem (M-TSP). In task sharing system, task balancing is made according to passive or active states. In load balancing, the goal is to avoid overloading a robot. After assignment of tasks to the relevant robots, the target cluster appears as many as the number of robots. For each set, the robot position and the available targets are considered as one of the graph nodes. The distance matrix is created by making these formed nodes as fully connected. Then, the path plan is made based on the proximity cost to the target nodes from the initial position of the robot (which is the starting node). When the next node to be moved is considered as the new starting position, each node that is visited, it is extracted from the graph connectivity matrix. The target and robots are labeled with colored labels and the positions of the objects are calculated by color-based quantization and thresholding methods. It has been observed that the system can make the task sharing and creates the appropriate path plan successfully with the variable target number and the different target distributions.Öğe Robot Control with Graph Based Edge Measure in Real Time Image Frames(Ieee, 2016) Donmez, Emrah; Kocamaz, Adnan Fatih; Dirik, MahmutRobotic path plan extraction is one of the major study area focused with image processing techniques except from estimation based methods, in real time robotic systems. In this study; I. First phase: It is aimed to track a mobile robot and target point by detecting start and target points in real time image frames, continually. II. Second phase: It is aimed to exhibit a novel sensor free kinematic model to determine power ratio transferred to the left and right wheels of the robot with decision tree method by utilizing a graph based method on detected object points. Right and left wheel position, robot center and labeled target positions are acquired with thresholding method by using labels placed on robot. A graph has been created on an image by admitting all detected areas as nodes. The processes of orientating and delivering mobile robot to the target position has been modelled according to distance values of edges between wheels and target.Öğe Static Path Planning Based on Visual Servoing via Fuzzy Logic(Ieee, 2017) Dirik, Mahmut; Kocamaz, Adnan Fatih; Donmez, EmrahRobot path planning algorithms and vision based solutions have gained importance among the algorithms developed to realize path planning in real time. In this paper, a new vision control model was developed for differential drive robot, using Fuzzy Logic decision sets that are independent of wheel encoder sensors. Real-time robot tracking and navigation was done using colored labels. A kinematic model with a graph based virtual sensor is utilized to develop a different solution to robot control problem.Öğe Vision-Based Decision Tree Controller Design Method Sensorless Application by using Angle Knowledge(Ieee, 2016) Dirik, Mahmut; Kocamaz, Adnan Fatih; Donmez, EmrahIn this study, a new control model for differential drive mobile robots was developed by using image-based decision tree method(DTM). Developed mobile robot control model was designed in an obstacle-free environment. The wheel encoder sensor was designed as a controller capable of independent positioning by using real-time images from overhead cameras on the bird's eye view. In this new method, a virtual triangular area between the target and the robot was created. These triangular base angles were calculated on the image. Decision tree controller was determined as the difference between the base angles by branching. Decision Tree leaves control determines the left and right wheel speeds depending on the difference model design was carried out. The developed new controller model was tested on Khepera IV robot. In practice, the robot's speed and angle of the body was carried uncensored control and it was observed to find the target in different applications. Application of the results and performance of the system was shown.Öğe A Vision-Based Real-Time Mobile Robot Controller Design Based on Gaussian Function for Indoor Environment(Springer Heidelberg, 2018) Donmez, Emrah; Kocamaz, Adnan Fatih; Dirik, MahmutIn this study, a visual servoing go-to-goal behavior controller is designed to control a differential drive mobile robot for a static target. Inputs for the controller method are based on a weighted graph or a triangle trigonometry kinematic model. The controller is designed with general Gaussian function by adapting the differential drive mobile robot dynamics. State parameters of dynamics are obtained by processing images in real time. It is aimed to develop an efficient internal sensor-independent visual-based control method. The single-head camera takes image frames from indoor environment. A real-time tracking process tracks the robot and target in sequential frames. The distances between graph nodes or the angles between edges are assigned as main control inputs according to utilized kinematic model. The velocity of wheels is computed for both models by using the general Gaussian function. We compare our method with two classical control methods that are PID and fuzzy-PID. Control of mobile robot has been made with high accuracy by using the designed visual-based controller.Öğe Visual Based Path Planning with Adaptive Artificial Potential Field(Ieee, 2017) Donmez, Emrah; Kocamaz, Adnan Fatih; Dirik, MahmutIn this study, extracting a path plan has been aimed by using vision based and potential field methods together for a static obstacle hosted environment. Path extracting operations have been executed on image frame taken from real environment with eye-out-device configuration. An obstacle free path plan is uncovered between robot and target with adaptive potential field method by considering position of robot, target and obstacle which are detected by image process techniques. Updating process of functional parameters belonging to potential field has been ensured with proposed method by taking into account distances of robot sensors to obstacle and global positions of obstacles. As a result of this study, strengths of the proposed method have been put forth by comparing experimental results of adaptive and non adaptive versions.Öğe Visual servoing based control methods for nonholonomic mobile robot(Academic Publication Council, 2020) Dirik, Mahmut; Kocamaz, Adnan Fatih; Donmez, EmrahIn this paper, we utilized two different vision-based go-to-goal robot control approaches on indoor nonholonomic mobile robot systems. In the proposed methods, eye-out-device configured camera (overhead camera) image data are used as the input parameters to determine the speeds of robot wheels. The main purpose of this system is to minimize the complexity of conventional robot control kinematics and to provide an efficient control approach to manage the wheel speeds and the direction angle of the mobile robot. In addition to reducing the complexity of robot control kinematics, it is also intended to reduce systematic and nonsystematic errors. The proposed method is divided into three stages: the first stage consists of the overhead camera calibration and the configuration of the robot motion environment. At this stage. the labels placed on the robot and target position were identified and the position information of the robot was obtained. In the second stage, control inputs such as position and orientation based on robot motion tracking and visual feature information were obtained. In the third stage, Graph-based Gaussian and Angle-based Decision tree control approaches were performed. We have briefly described these control approaches as follows: Graph-based Decision Tree Control (GDTC), Graph-based Gaussian Control (GGC), Angle-based Decision Tree Control (ADTC), and Angle-based Gaussian Control (AGC). Using these control approaches, many real-time experimental studies with eye-catching device configuration have been performed. The efficacy and usability of the methods have been demonstrated by experimental results.Öğe Visual Servoing Based Path Planning For Wheeled Mobile Robot in Obstacle Environments(Ieee, 2017) Dirik, Mahmut; Kocamaz, Adnan Fatih; Donmez, EmrahThe presentation of the environment based on visual data is important for the mobile robot to do localization and orientation. Selection safe path for the path planning, it is necessary to locate the position of the mobile robot in surrounding environment. In this study, vision-based control system and fuzzy logic controller methods were used together to construct a collision free path environment. The experimental environment was monitored with an overhead camera and robot position information, obstacles and target positions were determined by visual processing techniques. A safe (non-collision) path plan between the robot and the target has been achieved using fuzzy decision sets. Six virtual sensor data were used to plan the robot orientation control. A graphical representation of the test results of applications made for different scenarios is demonstrated and commentated.
