insufficient to correlate the images to other composed of four main - - PDF document

1 FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575”





Abstract— Imaging analysis is used for analyzing Frascati Tokamak Upgrade plasma. The overall data are organized in an open source portable software system based on a suitable database customized for the FTU. Mathematical and statistical models have been both developed in

• rder to correlate the images with plasma signals

and retrieve image information. Some Scientific Computing kernels

• f

the software are implemented in Problem Solving Environments Matlab that provides a comprehensive set of reference-standard algorithms and graphical tools for image processing and analysis. Using Matlab tools the characteristic components of plasma images have been observed.

Index Te1ms— Imaging analysis, Tokamak, Matlab tools, MySql.-Database, ENEA-GRID facilities.



I.INTRODUCTION

The advent of the internet and extensive digital image libraries have entailed the development

• f rapid and efficient computer-based image

searching and browsing techniques. The term content-based image retrieval refers to automatic recovery of images from a database based on a set of graphic features that qualify the images and that are, loosely speaking, similar to the characteristics of a given query

• image. This paper (within the CRESCO

Project) concerns the application of retrieval techniques and analysis of plasma images coming from Frascati Tokamak Upgrade (FTU). Data from FTU are acquired as movies

• r single frames. The wide number of

information available by visual inspection of the data is often a limitation to real time experimental investigation. It appears that the images from the cameras can help to reconstruct some plasma phenomena and real time analysis could be useful for machine

• peration. On the other hand, due to the large

number of data recorded, visual inspection of the movies is time consuming and often

Numerical and Statistical tools for images analysis based on the database from Frascati Tokamak Upgrade

• M. Chinnicia, S. Cuomob, S. Miglioric

a ENEA- FIM-INFOPPQ, Casaccia Research Center, S.Maria di Galeria (Roma), Italy

marta.chinnici@enea

bUNIVERSITA’ FEDERICO II, Dipartimento di Matematica e Applicazioni, Napoli , Italy cENEA- FIM, Enea Sede, Lungotevere Thaon di Revel n. 76 - 00196 Roma, Italy

1 FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575”

insufficient to correlate the images to other experimental findings. For a profitable use of the signals it is desirable to build specific tools for automatic processing of the images. The problems increase in complexity and cost in terms of hardware and software resources. In this research field the growing demand for computational resources has been a persistent goal of computational scientists. The challenge is to solve these problems with a computational power that could be achieved inexpensively by collecting “distributed” resources. The main goal is to provide both efficient and low cost computing environment by sharing the computational tools. This paper illustrates the acquisition system and the procedures developed for the processing and analysis of FTU images using ENEA-GRID technologies and middleware. In details, data are stored in a suitable database customized for the FTU, to be adopted in mathematical and statistical models in order to correlate the images with plasma signals and retrieve image information. This contribution is organized as follows. In section 2 we introduce the scenario of FTU. In section 3 we present the “FTUsoftware” methodology and we discuss the results of several retrieval experiments using the “FTUsoftware”. Section 4 concludes the paper.. FTU Scenario Frascati Tokamak Upgrade (FTU) is a compact, high magnetic field tokamak

• experiment. The objectives of the machine are

studying plasma transport, plasma heating and current drive in presence of strong additional radio frequency (RF) heating, and studying plasma profile by means of pellet injection. Plasma discharge lasts about 1.5 s, and in standard operating mode there is an experiment every 20–25 min. The FTU machine—from a control and data acquisition point of view—is composed of four main subsystems: the machine itself (torus, load assembly, cryostat, cooling system); the power supply (two fly- wheel generators, feeding the toroidal magnet and the poloidal windings respectively); RF facility and the diagnostic devices installed on the machine. The RF subsystem is made up of a lower hybrid system (LH: six gyrotrons

• perating at 8 GHz, total coupled power 2.5

MW), Ion–Bernstein Waves (IBW: three klystrons at 433 MHz, total coupled power 0.4 MW) and an electron cyclotron resonant heating system (ECRH: four gyrotrons at 140 GHz, total power 1.6 MW). The diagnostic devices —generally referred to as ‘diagnostics’—include detectors covering the whole range of the electromagnetic spectrum, particle density control devices like DCN laser,

• r experiments like multi-pellet injection.

In this paper, we focus our attention on images of plasma which are observed by wide angle video-cameras placed inside the viewing ports and close to the plasma edge. The cameras monitor the status of the vacuum vessel and toroidal limiter and give useful data

• n the occurrence of several events such as

arching and flying debris or even major damages such as the detachment of limiter

• plates. In details, the ports that we considered

are: Port_3_hor, Port_5_hor, Port_8_hor (where the suffix “hor” indicates horizontal video-cameras). “FTUsoftware” methodology In modern tokamaks visible and infrared video-cameras are becoming more and more important to monitor plasma evolution during fusion experiments. In the last years video- cameras have been extensively used in magnetic confinement fusion experiments for both the understanding of the physics and the safety of the operation. Both visible and Infra Red (IR) images can be used not only to

1 FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575”

monitor the evolution of a plasma discharge but also to evaluate specific parameters, from the determination of impurity radiation to the distribution of power loads on the plasma facing components. Data analysis is normally performed off-line, due to the high amount of information to be processed, making the data acquired by the camera quantitatively useful

• nly for post pulse evaluations. The main

difficulty in using visible or infrared images for plasma feedback control is the fact that real-time image processing is challenging and heavy in terms of processing time, especially when complex tasks are required. In order to manage, classify and process the images coming from FTU, we developed a software (“FTUsoftware”) that organizes the images in a database and allows to assess the information within the images. FTUsoftware is implemented by following the software engineering methodology. The main goal is to provide an integrated set of numerical and statistical tools and methods in a framework that is used to structure, plan and control the process of FTU data analysis. The software is developed in C++ language and the graphic interface is based on Open Qt libraries. The scientific computing kernel of the software is carried out on Problem Solving Environments Matlab that provides a comprehensive set

• f

reference-standard algorithms and graphical tools for image processing and analysis. The overall data of the system are stored in a DataBase (named tokamak) that is built in MySql (Fig.1)

• Fig. 1 Scheme of “Tokamak” DataBase.

In this paper we present a platform-based architecture for developing portable software in order to run it on any machine and any

• perating system without restrictions.

FTUsoftware usage and ENEA-GRID Middleware At the beginning it is necessary to set up the Middleware and install the appropriate libraries required to run the executable. The first step for accessing to images DataBase consists in inserting the IP numerical (e.g. 192.168.1.1) or string (e.g. “localhost”) data. Apart from the IP, the access requires to insert the name of the database, user and password; only authorized users can access to the platform. In the second step the system checks the existence of the file “config.conf” that gives us the information about the path of the images (Fig. 2).

1 FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575”

• Fig. 2 System’s pattern.

The presence of the file “Config.conf” guarantees that the FTU images DataBase exists, thus it not necessary to find the images inside the server. On the contrary, if we do not find this file, it is obligatory to insert the path

• f the server (Fig. 3) – path or machine’s

folders.

• Fig. 3 Using the path to find the images.

Step three. We examine the path; in fact, the path must be structured in a way as showed in

• Fig. 4.

.

• Fig. 4 Prefixed structure of software

system. Step four. FTUsoftware creates folders of

• ur data and if the server has been correctly

detected then it fills the folders recursively. Afterwards, FTUsoftware writes the “config.conf” file and the “treeDb.txt” file; the latter arranges the images of Database in order to create dynamical trees. After these phases, we may enter in the main page (Fig. 5) where we can operate on images. Fig 5 Operating window.

1 FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575”

The main page is divided in boxes; the first box on the left includes two sub-folders - Directory Server and Database. The first one contains the tree of the folders (inside the server) from which we pick up one or more images for analysis. The box has been set in order to show the last image under analysis, even if more images have been utilized.The second sub-folder – Database- shows the tree of the DataBase. At this point, the selected image is displayed in the box “Immagine originale”. When processed (e.g. gray-scale, restoration, segmentation)the image appears in the corresponding box (e.g. “Scala di grigi” etc.). (Fig. 6).

• Fig. 6 The boxes show the images under

analysis. It is possible to press the “Dimensioni

• riginali” key to visualize the original

dimension of the images (Fig. 7).

• Fig. 7 Original dimensions of an image.

It is possible to insert in or delete from the database an image by pressing “Add” or “Del”

• keys. Obviously, in order to avoid error, the

“Add” key will be active only if the image is not stored in the database, likewise the “Del” key will be active only if the image is already in the database. In Fig. 5-6-7, the first box on the right (Fig. 8) gives us the information about the selected image.

• Fig. 8 Information on image-data.

II.CONCLUSION This paper presents a software facility “FTUsoftware” aimed at analysing FTU images. The target is to use FTUsoftware in order to automate the processing of FTU images. The system is integrated in ENEA-GRID infrastructure and it is capable to link Numerical Mathematics with Statistical tools. FTUsoftware is an efficient, extremely robust

1 FINAL WORKSHOP OF GRID PROJECTS “PON RICERCA 2000-2006, AVVISO 1575”

and portable software; it allows to draw out and to detect/analyse the images within a database. ACKNOWLEDGMENTS This work is parts of the CRESCO Project funded by the ENEA (www.cresco.enea.it). REFERENCES

[1] Gonzalez , Digital Image Processing Using Matlab. [2]

• S. Cuomo, La Trasformata di Wavelet Parallela 2D,

di un'immagine digitale, CPS-CNR Tech. Rep. TR- 2001 - 04, Febbraio 2001 [3]

• M. Chinnici, S. Migliori, R. De Angelis, S. Borioni,
• S. Pierattini, Image analysis of a nuclear plasma:

Frascati Tokamak Upgrade using IDL and ENEA- GRID technologies (poster) - E-Science conference, Nsaples 2008 [4]

• R. De Angelis, S. Migliori, S. Borioni, G. Bracco, S.

Pierattini, A. Perozziello, Analysis of images from videocameras in the FTU tokamak, Review of scientific instruments, Vol. 75, N. 10 [5] http://www.cresco.enea.it/LA1/cresco_ sp12_graf3d/ [6] http://tangentsoft.net/mysql++/