Search

Your search keyword '"Dragan, Dinu"' showing total 41 results
Start Over Author "Dragan, Dinu" Publication Year Range Last 50 years
41 results on '"Dragan, Dinu"'

1. Deep Learning Methods for Retinal Blood Vessel Segmentation: Evaluation on Images with Retinopathy of Prematurity

Author

Gojić, Gorana, Petrović, Veljko, Turović, Radovan, Dragan, Dinu, Oros, Ana, Gajić, Dušan, and Horvat, Nebojša

Subjects
Electrical Engineering and Systems Science - Image and Video Processing, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Automatic blood vessel segmentation from retinal images plays an important role in the diagnosis of many systemic and eye diseases, including retinopathy of prematurity. Current state-of-the-art research in blood vessel segmentation from retinal images is based on convolutional neural networks. The solutions proposed so far are trained and tested on images from a few available retinal blood vessel segmentation datasets, which might limit their performance when given an image with retinopathy of prematurity signs. In this paper, we evaluate the performance of three high-performing convolutional neural networks for retinal blood vessel segmentation in the context of blood vessel segmentation on retinopathy of prematurity retinal images. The main motive behind the study is to test if existing public datasets suffice to develop a high-performing predictor that could assist an ophthalmologist in retinopathy of prematurity diagnosis. To do so, we create a dataset consisting solely of retinopathy of prematurity images with retinal blood vessel annotations manually labeled by two observers, where one is the ophthalmologist experienced in retinopathy of prematurity treatment. Experimental results show that all three solutions have difficulties in detecting the retinal blood vessels of infants due to a lower contrast compared to images from public datasets as demonstrated by a significant drop in classification sensitivity. All three solutions segment alongside retinal also choroidal blood vessels which are not used to diagnose retinopathy of prematurity, but instead represent noise and are confused with retinal blood vessels. By visual and numerical observations, we observe that existing solutions for retinal blood vessel segmentation need improvement toward more detailed datasets or deeper models in order to assist the ophthalmologist in retinopathy of prematurity diagnosis.

Published
2023
Full Text
View/download PDF

2. Overview of Deep Learning Methods for Retinal Vessel Segmentation

Author

Gojić, Gorana, Kundačina, Ognjen, Mišković, Dragiša, and Dragan, Dinu

Subjects
Electrical Engineering and Systems Science - Image and Video Processing, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Methods for automated retinal vessel segmentation play an important role in the treatment and diagnosis of many eye and systemic diseases. With the fast development of deep learning methods, more and more retinal vessel segmentation methods are implemented as deep neural networks. In this paper, we provide a brief review of recent deep learning methods from highly influential journals and conferences. The review objectives are: (1) to assess the design characteristics of the latest methods, (2) to report and analyze quantitative values of performance evaluation metrics, and (3) to analyze the advantages and disadvantages of the recent solutions.

Published
2023

3. Non-adversarial Robustness of Deep Learning Methods for Computer Vision

Author

Gojić, Gorana, Vincan, Vladimir, Kundačina, Ognjen, Mišković, Dragiša, and Dragan, Dinu

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition
Abstract

Non-adversarial robustness, also known as natural robustness, is a property of deep learning models that enables them to maintain performance even when faced with distribution shifts caused by natural variations in data. However, achieving this property is challenging because it is difficult to predict in advance the types of distribution shifts that may occur. To address this challenge, researchers have proposed various approaches, some of which anticipate potential distribution shifts, while others utilize knowledge about the shifts that have already occurred to enhance model generalizability. In this paper, we present a brief overview of the most recent techniques for improving the robustness of computer vision methods, as well as a summary of commonly used robustness benchmark datasets for evaluating the model's performance under data distribution shifts. Finally, we examine the strengths and limitations of the approaches reviewed and identify general trends in deep learning robustness improvement for computer vision.

Published
2023

4. Overview of Deep Learning Methods for Retinal Vessel Segmentation

Author

Gojić, Gorana, Kundačina, Ognjen, Mišković, Dragiša, Dragan, Dinu, Kacprzyk, Janusz, Series Editor, Gomide, Fernando, Advisory Editor, Kaynak, Okyay, Advisory Editor, Liu, Derong, Advisory Editor, Pedrycz, Witold, Advisory Editor, Polycarpou, Marios M., Advisory Editor, Rudas, Imre J., Advisory Editor, Wang, Jun, Advisory Editor, Trajanovic, Miroslav, editor, Filipovic, Nenad, editor, and Zdravkovic, Milan, editor

Published
2024
Full Text
View/download PDF

6. Automatic corrections of human body depth maps using deep neural networks

Author

Gojić Gorana, Turović Radovan, Dragan Dinu, Gajić Dušan, and Petrović Veljko

Subjects
depth map refinement, human body, photogrammetry, deep learning, segmentation, Electrical engineering. Electronics. Nuclear engineering, TK1-9971
Abstract

This paper presents an approach to correcting misclassified pixels in depth maps representing parts of the human body. A misclassified pixel is a pixel of a depth map which, incorrectly, has the ‘background’ value and does not accurately reflect the distance from the sensor to the body being scanned. A completely automatic, deep learning based solution for depth map correction is proposed. As an input, the solution requires a color image and a corresponding erroneous depth map. The input color image is segmented using deep neural network for human body segmentation. The extracted segments are further used as guidance to find and amend the misclassified pixels on the depth map using a simple average based filter. Unlike other depth map refinement solutions, this paper designs a method for the improvement of the input depth map in terms of completeness instead of precision. The proposed method does not exclude the application of other refinement methods. Instead, it can be used as the first step in a depth map enhancement pipeline to determine approximate depths for erroneous pixels, while other refinement methods can be applied in a second step to improve the accuracy of the recovered depths.

Published
2020
Full Text
View/download PDF

9. Automation of the avatar animation process in FBX file format

Author

Lazor Milica, Gajić Dušan B., Dragan Dinu, and Duta Alina

Subjects
computer animation, computer graphics, automation, avatars, fbx, c++ programming, Engineering (General). Civil engineering (General), TA1-2040, Mechanics of engineering. Applied mechanics, TA349-359
Abstract

Animated three-dimensional (3D) models are an essential part of interactive content available in products of present day movie and gaming industries. Time required for manual development of animated 3D models is a limiting factor in many of their practical applications. In order to minimize the production time of animated 3D models and, thus, lower costs, there is a need for automation of the animation process. This paper presents an approach to the programmatic addition of animations to automatically rigged avatars stored using the industry-standard FBX file format. The presented solution is developed using the C++ programming language and the Autodesk FBX Software Development Kit (SDK). It allows significantly higher degree of automation and programmability than other standard approaches, based on using standard 3D modelling and animation software, such as Autodesk 3Ds Max or Maya.

Published
2019

21. Robustness of deep learning methods for ocular fundus segmentation: Evaluation of blur sensitivity.

Author

Petrović, Veljko B., Gojić, Gorana, Dragan, Dinu, Gajić, Dušan B., Horvat, Nebojša, Turović, Radovan, and Oros, Ana

Subjects
FUNDUS oculi, RANDOM noise theory, NOISE control, FAILURE mode & effects analysis, DEEP learning, CONVOLUTIONAL neural networks
Abstract

This article analyzes the sensitivity of deep learning methods for ocular fundus segmentation. We use an empirical methodology based on non‐adversarial perturbed datasets. The research is motivated by mass screening and self‐administered tests in which AI methods are needed and may be given images with focus issues. These substandard pictures are simulated using blurring algorithms of varying designs and kernel sizes which are subjected to a test of inter‐network and inter‐dataset sensitivity using Gaussian noise as a control. We test for simulated defocusing, motion blur, and Gaussian noise. The DRIVE, CHASE, and STARE datasets were used to generate 441 synthetic datasets for testing. Analysis of the resultant n = 37,888 sample has identified anomalies illuminating failure modes of state‐of‐the‐art ocular segmentation models. Data analysis led us to conclude that accuracy is a poor measure of sensitivity to input change and that even moderate levels of blur have dramatic effects on said sensitivity. We further concluded that architectures show larger variations in sensitivity to blur which do not match either their reported or measured performance on unblurred test datasets. Our analysis has attributed at least a part of the problem to overfitting non‐essential input dataset features and resolution sensitivity. [ABSTRACT FROM AUTHOR]

Published
2022
Full Text
View/download PDF

22. Програмски оквир заснован на машинском учењу за аутоматизацију обраде резултата фотоакустичних мерења

Author

Kupusinac, Aleksandar, Ivetić, Dragan, Popović, Marica, Gajić, Dušan, Dragan, Dinu, Јордовић-Павловић, Мирослава, Kupusinac, Aleksandar, Ivetić, Dragan, Popović, Marica, Gajić, Dušan, Dragan, Dinu, and Јордовић-Павловић, Мирослава

Abstract

Главни задатак истраживања приказаног у дисертацији је развој модела, заснованог на алгоритмима машинског учења, који описује сложени утицај мерног система на користан, експериментални сигнал са циљем његове елиминације. Студија случаја је широко распрострањена фотоакустична, трансмисиона мерна метода са ћелијом минималне запремине. Мултидисциплинарност и комплексност проблема одредили су следеће кораке у методологији решења: 1) развој софтвера за генерисање симулираних експерименталних података, 2) развој регресионог модела заснованог на трослојној неуронској мрежи, за прецизну и поуздану карактеризацију детектора која се извршава у реалном времену, 3) развој класификационог модела заснованог на неуронској мрежи једноставне структуре за прецизну и поуздану предикцију типа коришћеног детектора која се извршава у реалном времену, 4) спрезање регресионог и класификационог модела уз развој додатног софтвера за прилагођење модела стварном експерименту. На овај начин заокружен је програмски оквир који извршава сложени задатак издвајања “правог” сигнала oд изобличеног експерименталног сигнала без ангажовања истраживача, односно извршава аутокорекцију. Тестирање је извршено на више различитих детектора и више различитих материјала у фотоаксустичном експерименту. Применом развијеног програмског оквира конкурентност експерименталне технике је знатно порасла: повећана је тачност и поузданост, проширен је мерни опсег и смањено време обраде резултата мерења., Glavni zadatak istraživanja prikazanog u disertaciji je razvoj modela, zasnovanog na algoritmima mašinskog učenja, koji opisuje složeni uticaj mernog sistema na koristan, eksperimentalni signal sa ciljem njegove eliminacije. Studija slučaja je široko rasprostranjena fotoakustična, transmisiona merna metoda sa ćelijom minimalne zapremine. Multidisciplinarnost i kompleksnost problema odredili su sledeće korake u metodologiji rešenja: 1) razvoj softvera za generisanje simuliranih eksperimentalnih podataka, 2) razvoj regresionog modela zasnovanog na troslojnoj neuronskoj mreži, za preciznu i pouzdanu karakterizaciju detektora koja se izvršava u realnom vremenu, 3) razvoj klasifikacionog modela zasnovanog na neuronskoj mreži jednostavne strukture za preciznu i pouzdanu predikciju tipa korišćenog detektora koja se izvršava u realnom vremenu, 4) sprezanje regresionog i klasifikacionog modela uz razvoj dodatnog softvera za prilagođenje modela stvarnom eksperimentu. Na ovaj način zaokružen je programski okvir koji izvršava složeni zadatak izdvajanja “pravog” signala od izobličenog eksperimentalnog signala bez angažovanja istraživača, odnosno izvršava autokorekciju. Testiranje je izvršeno na više različitih detektora i više različitih materijala u fotoaksustičnom eksperimentu. Primenom razvijenog programskog okvira konkurentnost eksperimentalne tehnike je znatno porasla: povećana je tačnost i pouzdanost, proširen je merni opseg i smanjeno vreme obrade rezultata merenja., The main task of the research presented in this dissertation is the development of the model based on machine learning algorithms, which describes the complex influence of the measuring system on a useful, experimental signal, with the aim of the elimination of this influence. The case study is a widespread photoacoustic, transmission measurement method with minimum volume cell configuration. Multidisciplinarity and complexity of the problem determined the following steps in the solution methodology: 1) development of the software for generating simulated experimental data, 2) development of the regression model based on a three-layer neural network, for precise and reliable characterization of detectors, performed in real time, 3) development of the classification model based on a neural network of simple structure for precise and reliable prediction of the type of detector in use, performed in real time, 4) coupling of the regression and the classification model with the development of additional software for adjustment of the model to a real experiment. In this way, the program framework is completed, which performs the complex task of extracting the "true" signal from the distorted experimental signal without the involvement of researchers, performing, thus, the autocorrection. Testing was performed on several different detectors and several different materials in a photoacoustic experiment. With the application of the developed software framework, the competitiveness of the experimental technique has significantly increased: the accuracy and the reliability have been increased, the measurement range has been expanded and the processing time of measurement results has been reduced.

Published
2020

29. Architectures of DICOM based PACS for JPEG2000 medical image streaming

Author

Dragan Ivetic and Dragan Dinu

Subjects
Ubiquitous computing, General Computer Science, Multimedia, Syntax (programming languages), business.industry, Computer science, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Data_CODINGANDINFORMATIONTHEORY, computer.file_format, computer.software_genre, Task (computing), DICOM, Picture archiving and communication system, Transmission (telecommunications), JPEG 2000, business, computer, Computer hardware, Image compression
Abstract

Delivering of medical image content to mobile/embedded devices with low storage and processing capabilities and low resolution displays is a challenging, but important task for achieving ubiquitous computing in modern hospital environments. The common approaches in industry and technical literature employ JPEG2000 compression and image streaming. Beside image compression, JPEG2000 facilitates image streaming. Different sized images are extracted from one codestream with minimal spatial distortion within decompressed image. We developed DICOM2000 syntax which enables JPEG2000 streaming over DICOM networks. DICOM2000 brings compression and streaming power of JPEG2000 to DICOM which is the standard for medical image interchange. It enables transmission of best quality medical images suited for client device processing and display capabilities. In this paper communication architectures of DICOM2000 based PACS are compared with the common architectures of DICOM based PACS. For the purpose of the paper, these communication architectures are categorized, named, and described. At the end, the advantages of the DICOM2000 based PACS over standard PACS are pointed up.

Published
2009

30. DICOM Overview

Author

Dragan Dinu and Dragan Ivetic

Subjects
lcsh:TA1-2040, lcsh:Engineering (General). Civil engineering (General)
Abstract

The paper introduces preliminary results of our research on DICOM – JPEG 2000 coupling: DICOM standard. DICOM standard defines method for transferring digital images of various formats and associated data between devices manufactured by various vendors. PACS (Picture Archiving and Communication System) inside a hospital that can interface with other informational subsystems is based on this standard. DICOM standard is applicable to a network and/or to an off-line media environment. These features of DICOM standard manage communication between PACS in different hospitals. That means that if patient is transferred to another hospital his new physician can access patient’s medical records over the network. They also manage online accession to medical records from physician’s home. The standard supports operations that are based on other computer protocols (TCP/IP, ISO 9660, etc). Digital image and associated data are coupled in a single DICOM file. Image pixels are encapsulated inside this file. Format of image can be native DICOM format and/or some other lossless or lossy standard (jpeg, jpeg 2000, etc). At the end, development of PACS system of Hospital for chest-diseases which is based on DICOM standard is discussedi.

Published
2005

32. CREATION AND USAGE OF 3D FULL BODY AVATARS.

Author

BERDIC, Nikola, DRAGAN, Dinu, MIHIC, Srdan, and ANISIC, Zoran

Subjects
*AVATARS (Virtual reality), *3-D animation, *CONSUMER goods, *VIRTUAL reality software, *COMPUTER graphics
Abstract

Different ways of 3D avatar creation and their current everyday use is described in the paper together with discussion on expansion of 3D avatar use in the future. 3D avatar is the complete, customized, and animated graphical representation of a person in three-dimensions. 3D full body avatar is a special type of 3D avatar as it represents a digital copy of person's full body. 3D avatars come in many different forms. They differ in level of details, realism (how closely is the person represented with 3D avatar), and behavior (what moves and mimics the avatar supports). Different techniques, such as manual and/or automated software 3D modeling, or sophisticated 3D scanning techniques, are employed in creating 3D avatars. As our society becomes more and more digitalized, the amount of data regarding one person becomes enormous, and what better way of representing that person's data then using the person's avatar. [ABSTRACT FROM AUTHOR]

Published
2017

33. REVIEW OF 3D BODY SCANNING SYSTEMS.

Author

ARBUTINA, Miroljub, DRAGAN, Dinu, MIHIC, Srdan, and ANISIC, Zoran

Subjects
*SCANNING systems, *THREE-dimensional imaging, *DIGITIZATION, *COMPUTER graphics, *PHOTOGRAMMETRY
Abstract

The purpose of this article is to review 3D body scanning systems currently available in the market. 3D human scanning is a process of capturing and subsequent digitization of the full body or the body parts into their three-dimensional graphical representation. As 3D human body scanning techniques are becoming ubiquitous, we give an overview of the different technologies of 3D body data digitization such as techniques based on photogrammetry, depth sensors, and laser scanning as well as few more promising techniques. Also, the underlying principles that allow these systems to work are determined and explained. In order to provide directions for further integration of digitized 3D human bodies, specifications of 3D body scanning systems are compared, and a variety of commercial scanners available in the market are presented in the paper. [ABSTRACT FROM AUTHOR]

Published
2017

38. Model upravljanja velikim serijama geoprostornih podataka

Author

Pajić, Vladimir, Govedarica, Mitar, Ristić, Aleksandar, Stojanović, Dragan, Jovanović, Dušan, Dragan, Dinu, and Dinu, Dragan

Subjects
Big Data, Geospatial data, point cloud, space-filling curves, ApacheSpark, Big Data, geoprostorni podaci, oblak tačaka, krive za popunjavanjeprostora, Apache Spark
Abstract

U okviru disertacije je predstavljen je model za upravljanje velikimserijama geoprostornih podataka baziran na Spark platformi. Model jefokusiran na oblak tačaka, s obzirom na njegove specifičnekarakteristike. Posebno je razmatrano indeksiranje geoprostornihpodataka u distribuiranom okruženju bazirano na krivama zapopunjavanje prostora. Model podržava osnovne fukcionalnosti poput,strukturiranja i skladištenja podataka, vizualizacije, i geoprostornihanaliza. Izvršena je verifikacija kroz niz eksperimenata, uključujućipoređenje sa relacionim bazama podataka, i ispitivanje skalabilnosti., The thesis presents a large series of geospatial data management modelbased on Spark framework. Model is focused on the point cloudgeospatial data type due to it’s specific characteristics. Indexing ofgeospatial data in distributed environment based on space-fillingcurves is described in detail. The proposed model provides completeset of functionalities like structuring and storing the data, visualizationand geospatial analyses. The model is verified through the set ofexperiments, including comparison to relational databases andscalability assesment.

Published
2021

39. Programski okvir zasnovan na mašinskom učenju za automatizaciju obrade rezultata fotoakustičnih merenja

Author

Jordović, Pavlović Miroslava, Kupusinac, Aleksandar, Ivetić, Dragan, Popović, Marica, Gajić, Dušan, and Dragan, Dinu

Subjects
Mašinsko učenje, Veštačke neuronske mreže, Regresija, Klasifikacija, Analiza glavnih komponenti, Fotoakustika, Mikrofon, Машинско учење, Вештачке неуронске мреже, Регресија, Класификација, Анализа главних компоненти, Фотоакустика, Микрофон, Machine Learning, Artificial Neural Networks, Regression, Classification, Principal Component Analysis, Photoacoustics, Microphone
Abstract

Главни задатак истраживања приказаног у дисертацији је развој модела,заснованог на алгоритмима машинског учења, који описује сложениутицај мерног система на користан, експериментални сигнал са циљемњегове елиминације. Студија случаја је широко распрострањенафотоакустична, трансмисиона мерна метода са ћелијом минималнезапремине. Мултидисциплинарност и комплексност проблема одредилису следеће кораке у методологији решења: 1) развој софтвера загенерисање симулираних експерименталних података, 2) развојрегресионог модела заснованог на трослојној неуронској мрежи, запрецизну и поуздану карактеризацију детектора која се извршава уреалном времену, 3) развој класификационог модела заснованог нанеуронској мрежи једноставне структуре за прецизну и поузданупредикцију типа коришћеног детектора која се извршава у реалномвремену, 4) спрезање регресионог и класификационог модела уз развојдодатног софтвера за прилагођење модела стварном експерименту. Наовај начин заокружен је програмски оквир који извршава сложени задатакиздвајања “правог” сигнала oд изобличеног експерименталног сигналабез ангажовања истраживача, односно извршава аутокорекцију.Тестирање је извршено на више различитих детектора и вишеразличитих материјала у фотоаксустичном експерименту. Применомразвијеног програмског оквира конкурентност експерименталне техникеје знатно порасла: повећана је тачност и поузданост, проширен је мерниопсег и смањено време обраде резултата мерења., Glavni zadatak istraživanja prikazanog u disertaciji je razvoj modela,zasnovanog na algoritmima mašinskog učenja, koji opisuje složeniuticaj mernog sistema na koristan, eksperimentalni signal sa ciljemnjegove eliminacije. Studija slučaja je široko rasprostranjenafotoakustična, transmisiona merna metoda sa ćelijom minimalnezapremine. Multidisciplinarnost i kompleksnost problema odredilisu sledeće korake u metodologiji rešenja: 1) razvoj softvera zagenerisanje simuliranih eksperimentalnih podataka, 2) razvojregresionog modela zasnovanog na troslojnoj neuronskoj mreži, zapreciznu i pouzdanu karakterizaciju detektora koja se izvršava urealnom vremenu, 3) razvoj klasifikacionog modela zasnovanog naneuronskoj mreži jednostavne strukture za preciznu i pouzdanupredikciju tipa korišćenog detektora koja se izvršava u realnomvremenu, 4) sprezanje regresionog i klasifikacionog modela uz razvojdodatnog softvera za prilagođenje modela stvarnom eksperimentu. Naovaj način zaokružen je programski okvir koji izvršava složeni zadatakizdvajanja “pravog” signala od izobličenog eksperimentalnog signalabez angažovanja istraživača, odnosno izvršava autokorekciju.Testiranje je izvršeno na više različitih detektora i višerazličitih materijala u fotoaksustičnom eksperimentu. Primenomrazvijenog programskog okvira konkurentnost eksperimentalne tehnikeje znatno porasla: povećana je tačnost i pouzdanost, proširen je merniopseg i smanjeno vreme obrade rezultata merenja., The main task of the research presented in this dissertation is the developmentof the model based on machine learning algorithms, which describes thecomplex influence of the measuring system on a useful, experimental signal,with the aim of the elimination of this influence. The case study is a widespreadphotoacoustic, transmission measurement method with minimum volume cellconfiguration. Multidisciplinarity and complexity of the problem determined thefollowing steps in the solution methodology: 1) development of the software forgenerating simulated experimental data, 2) development of the regressionmodel based on a three-layer neural network, for precise and reliablecharacterization of detectors, performed in real time, 3) development of theclassification model based on a neural network of simple structure for preciseand reliable prediction of the type of detector in use, performed in real time, 4)coupling of the regression and the classification model with the developmentof additional software for adjustment of the model to a real experiment. In thisway, the program framework is completed, which performs the complex taskof extracting the "true" signal from the distorted experimental signal without theinvolvement of researchers, performing, thus, the autocorrection. Testing wasperformed on several different detectors and several different materials in aphotoacoustic experiment. With the application of the developed softwareframework, the competitiveness of the experimental technique hassignificantly increased: the accuracy and the reliability have been increased,the measurement range has been expanded and the processing time ofmeasurement results has been reduced.

Published
2020

40. Pristupi razvoju baze podataka Opšteg informacionog modela za elektroenergetske mreže

Author

Dević, Saša, Luković, Ivan, Švenda, Goran, Tomašević, Milo, Kordić, Slavica, and Dragan, Dinu

Subjects
Common Information Model (CIM), database modeling, code generators, power grid, model to code transformation, Opšti informacioni model (CIM), modelovanje baze podataka, generisanje koda, elektroenergetske mreže, model-kod transformacije, Општи информациони модел (CIM), моделовање базе података, генерисање кода, електроенергетске мреже, модел-код трансформације
Abstract

Општи информациони модел (CIM) користи се за опис електроенергетске мреже и за размену података између оператера преносних електроенергетских система. Како је модел постајао све заступљенији, појавила се потреба за његовим складиштењем. У раду је развијен методолошки приступ за развој базе података која би подржала релативно једноставно складиштење и рад са инстанцама CIM модела, које описују тренутно, активно стање у систему. Такође, омогућено је и праћење претходних, историјских стања CIM инстанци, као и њихова рестаурација у жељено стање. Очекује се да предложени приступ олакша увођење CIM модела у различита, наменска програмска решења., Opšti informacioni model (CIM) koristi se za opis elektroenergetske mreže i za razmenu podataka između operatera prenosnih elektroenergetskih sistema. Kako je model postajao sve zastupljeniji, pojavila se potreba za njegovim skladištenjem. U radu je razvijen metodološki pristup za razvoj baze podataka koja bi podržala relativno jednostavno skladištenje i rad sa instancama CIM modela, koje opisuju trenutno, aktivno stanje u sistemu. Takođe, omogućeno je i praćenje prethodnih, istorijskih stanja CIM instanci, kao i njihova restauracija u željeno stanje. Očekuje se da predloženi pristup olakša uvođenje CIM modela u različita, namenska programska rešenja., Common Information Model (CIM) is used for describing power grid networksand data exchange among transmission system operators (TSO). As themodel became widely used, there was a need to store such model. In thisthesis we present a methodological approach to development of a databasethat supports relatively easy storing and managing CIM instances, whichdescribe current, active state of the system. Also, tracking changes andrestoring CIM instances to its previous states are supported. We expect thatsuch methodological approach would ease the implementation of CIM modelin various, domain specific software solutions.

Published
2019

41. Namenski jezik za vizuelizaciju evaluiranu statističkom analizom malih skupova podataka

Author

Petrović, Veljko, Živanov, Žarko, Gilezan, Silvia, Dragan, Dinu, Rančić, Dejan, and Grgić, Zorka

Subjects
vizuelizacija, Domain Specific Languages (DSL), visual handicaps, statistička analiza malih skupova podataka, Namenski jezici, Namenski jezici, vizuelizacija, poremećaji vida, statistička analiza malih skupova podataka, evaluacija vizuelizacija, Domain Specific Languages (DSL), visuelziation, visual handicaps, statisticalanalysis of small data sets, visualization evaluation, evaluacija vizuelizacija, poremećaji vida, visualization evaluation, visuelziation, statisticalanalysis of small data sets
Abstract

Za potrebe evaluacije kvaliteta vizuelizacije u svrhu prakse inkluzivnog dizajna razvijen je namenski jezik koji hendikepe vida opisuje kroz formalizam baziran na vizuelnim promenljivama. Namenski jezik se može koristiti kao dokumentacija, ili kao specifikacija hendikepa za posebno napisan simulator vizuelnih hendikepa koji se može koristiti za potrebe testiranja. Kao pandan ovoj tehnologiji razvijen je novi mehanizam za evaluaciju vizuelizacija koji kroz inovacije u metodologiji i statistici omogućava da se merodavni zaključci izvedu iz relativno malih uzoraka podataka., For purposes of visualization quality evaluation as part of the practice ofinclusive design, a domain specific language was developed such that itdescribes visual handicaps through a formalism based on visual variables.This domain specific language can be used as documentation or as a handicapspecification for a specially constructed visual handicap simulator usable fortesting. Alongside this technology, a novel mechanism for visualizationevalution was developed. This mechanism allows, through innovations inmetholodoly and statistics, the production of trustworthy results from relativelysmall data sets.

Published
2018

Catalog

Books, media, physical & digital resources
See catalog results