CV SANDOR SERES
Personalities
Date of birth 8. September 1939
Nationality Norwegian
Contact Details E-mail:
Tlf: +47 92617447
Forwarding address Jarmyrveien 3, 1358 Jar, Norway.
Until 1985 I lived in the former Yugoslavia and worked in the university education and in the scientific research area. In 1985, by chance, my family moved to Norway. From that time I have been working on applied computer research and technology development, mostly in the Image Processing area.
I am married and have two grownup children.
Education
BSc, Mathematics.
Engineer of Electronics.
M.Sc, Mathematics, Theory of Stochastic systems and Information Theory.
University of Belgrade, former Yugoslavia.
PhD degree in Technical Sciences, Control of Stochastic Systems
University of Novi Sad, former Yugoslavia.
I use Norwegian, Hungarian, Serbo-Croatian and English languages fluently. Besides, I read Russian and German.
Scientific experience
Until 1985 I was working mostly on problems related to Complex Systems. My major concern was – how to replace the highly formal and narrow mathematical System Theory with more useful and less formal structural system analysis and design. In this context, my main focus was on systems’ – entropy sources and their control via information systems.
From that period, one of my highly theoretical researches was from the AI area. More precisely, it was about one of its basic problems – the General Recognition Problem. Namely, how to recognize in a large set of similar objects using their known attributes. This problem very quickly leads to the well-known Combinatorial Explosion in the classical case-comparison methods. I have developed an algorithm to create the so-called Optimal Convex Order of attributes. Also, I have proofed that there is no better combination of N attributes than the first N attributes from the Optimal Convex Order. These N attributes guarantee the most efficient recognition.
Finally, I was member of a large research team that has developed an Information System model for large social/complex systems like large towns, communities… states. I was working on the kernel of this modular and flexible system.
Parallel with these researches I was in many years involved into education processes at the University of Novi Sad. My last position was titled – University Professor what I still keep today. My subjects were: General System’s Theory, Information Theory and Information System Design Methodologies.
During this period I was author to several University books for the mentioned subjects as well as co-author for even more University books. Also, I wrote a monography book about the mentioned General Recognition Problem. Finally, I wrote also a number of articles and seminar-reports (many of these were necessary for formal qualifications in the University hierarchy).
Practical and professional experience
From the beginning of my working life I was very interested in practical work, to se the results of my effort. Therefore I have started my computer related studies pretty early. Interest for programming, geometry and iterative approaches on computers, especially Image Processing, became a kind of compulsion to me. Even today, these areas are my great hobbies and a kind of psychological medicine to me.
There are so many projects I was involved with that there is no place here for a full list and description of them. So, I will just mention some of these.
-Information system’s design and implementation in some large companies.
-Computer based Operational research, optimizations and simulations.
-Stock control system, inclusive forecast, design and implementation (in the largest copper-mining company in Europe and in an oil-products distribution company).
-Large format scanner development and raster based Image Processing. The size of these flat-bad scanners goes up to 8m*3m and the resolution up to 2032 dpi. Consequently, these scanners produce data at some GB level. So, the compression and the format handling were an alpha problem and my basic task.
-Different computer graphics oriented projects like: raster and vector image editors (kernel functions like non-pixel based layered bi-tonal raster image handing, geometry primitives rendering like: thick lines, circular/elliptical arcs, text…) , plotter drivers, transformations, raster-to-vector conversion, pattern recognition, archiving, position-images based location detection, collision control and avoidance, safety monitoring and so on.
-Network information system’s design and implementation (for large telecom companies like TeleNor, TeleDanmark…).
-Position images based Spatial Object Query methods, as a simple and highly efficient alternative for DBMS Spatial Options (fe. Oracle,s Spatial Option).
And so on.
Finally, I am/was very well acquainted with the historical Fortran, Algol and Cobol programming languages.
I am also familiar with some DBMS and script languages like TCL, COMs, PSQL, JAVA. Lately, I am predominantly using C++ and C in the digital mapping and image processing systems development.
Employment and positions from 1985
2011-present, free-lance and enthusiast OSM (Open Street Map) related development like: data-preparation-tool-chain, vector and hybrid data format, a new reed-only high efficient server and so on.
2002-2011, Faster Imaging AS, Oslo, Norway, founder and Chief Technical Officer.
Responsible for all development and platform independent applications coordination.
2001-2002 Norwegian Computing Center, Oslo, Norway, Chief Researcher.
Responsible for the Position Detection of the Cellular mobile devices.
1992-2000 Incatel AS Oslo, Norway/AnnArbour US, Principal Software Engineer. Responsible for the Base-Map applications within the NIMS Plant Management Information System.
1989-1992 Cimage Corporation Oslo, Norway/Copertino US, Basic-development department manager.
Responsible for the raster and vector graphics editor kernel development.
1985-1989 Eldak, Horten, Norway. System engineer.
Responsible for the raster-to-vector conversion project.
Current interests
My current professional interest and occupation is just a continuation of my former hobby Image Processing related activities. Shortly, these activities, and the present results, all are originally developed by me and may be grouped into five technology categories:
-Layered bi-tonal raster images technology. All functions are running directly on a highly compressed internal presentation with the “movie performance”. The number of layers and their sizes are practically unlimited. Because of my Image Algebra, the number of possible imaging functions is limited only with the fantasy.
-Vector/parametric defined image technology. Again, all functions that handle vector geometry are originally developed by me inclusive the smallest details like: area clipping (note, not polygon clipping), area fill, inside/outside area test, line-thickness continual scaling (down to sub-pixel level)… The visual impression, or quality, is rotation and scaling invariant. The rendering technology is my “smooth-edge” color-to-color rendering that provides excellent readability and allows many innovative functions.
-Font and text rendering technology. My fonts are “alike” Arial, Georgia, Greek, Tahoma…but presented with hundreds of times less data volume compared to the originals. The smooth-edge rendering provides excellent readability and full control of all character and text parameters.
-The aero-photos/picture-maps handling technology. My color-space reduction and format preparation innovation provides a movie performance in rendering and navigation over thousands of stitched together images.
-Finally, my vector mapping technology was (and is) my major hobby and interest in many years. The application/experimental arena may be any of the known World geo-data sets like TA, NT, AND… but from some years ago I am mostly using OSM Planet data. This technology comprises many unique innovative algorithms and programs for data-preparation (error detection and reduction, simplification, scale/zoom levels’ generation with data reduction, multi-tiling an some more), read-only server-data preparation, request handling and transmission format preparation. The mentioned server model is extraordinary efficient. It manages many thousands of requests/second (on a moderate HW), it is based on the multi-tiling (the final/small tiles are created on-the-fly and may have arbitrary size) the transmission format is highly compact (20 or more times less data compared to the commonly used PNG format).
It is also worth to mention the many valuable original tools for experimenting, viewing, analyzing, comparing…that were necessary during the development of the mentioned technologies.
Oslo, 25.08.13.
Sandor Seres
4