Liudmila Burțeva

Institute of Mathematics and Computer Science · System Programming Lab

This paper discusses the application of ABBYY FineReader for the recognition of three Romanian manuscripts written in un- cial Cyrillic script. The manuscripts, dating from the 16th cen- tury, were digitized and processed achieving an OCR accuracy between 75% and 85%. The study demonstrates the effective- ness of specialized OCR models in handling...

The presented work refers to implementation of Digital Science concept. The essential component of digitizing scientific research is the revitalization of existing printed publications. This process aims not only to provide access to publications by making them suitable for digital archives but also to enable the reuse of their content in new works...

This book examines the problems of digitization of both types of documents - homogeneous and heterogeneous, printed in the Roman language with Cyrillic characters. It is based on the publications of seven years of research and development by our team in this field, discusses the current state, challenges, and perspectives.

In this paper, we summarize the research of digitization of documents printed by Romanian transitional alphabet. These printings are the most original Romanian historical documents, which makes our experience useful when researching OCR methods for similar alphabets. The current work is focused to OCR that is the first stage of scanned documents di...

This chapter presents the technologies and approaches devel- oped at V. Andrunachievici Institute of Mathematics and Computer Science in Chisinau, Republic of Moldova, for digitization of old Romanian documents. It is an overview of ongoing research and development over the past seven years and reflects its current state.

The presented work concerns the current stage of the research of a purely P system solution of the problem of clustering of two-dimensional points. For this task, black and white images are mainly considered to be the most appropriate representation of the input data. Such image-based representation inspires the idea of presented work. In frame of...

In the presented work, the developed by author segmentation technique based on P system is applied to pre-OCR enhancing of document images. Developed for solving the noisy medical imaging problems, the proposed technique has the advantage features, which ensure successful application to the historical document images enhancement problem. The first...

The archives of scanned printed documents grow dramatically by results of projects concerning cultural heritage preserving. A manual annotations of scanned document images and per page screen reading make the usage of these achieves difficult and, sometimes, impossible. Existing document processing systems cannot automatically display content corre...

This article considers aspects of application of advanced parallel computing techniques to computational linguistics Big Data problems, which spring from creation Web-based corpora on base of digitization of huge collections of historic texts scans. Solving of the majority of linguistic research problems requires per word processing that is the nat...

This paper presents the advantages of P system computing based approach to solving image processing problems. Being sig- nificantly reduced relative to classic algorithms, P system com- puting based algorithms nevertheless produce better results and even resolve the problems just unsolvable for classic algorithms. Despite these evident advances, un...

This paper presents the approach for medical imaging problem solving by application of P systems which are branch of unconventional computing. Presented results show the advances of this approach. P system computing allows development of algorithms significantly reduced related to classic one. These new algorithms nevertheless produce better result...

This is an introductory paper of fresh started research that concerns the development of new HPC parallelization strategies based on intrinsically parallel unconventional computation paradigms. The main goal of this research is to use theses developed parallelization technologies for overcoming modern HPC energy consumption issues. Development of n...

This work presents a hybrid model of high performance computations, representing the P system framework with additional quantum functionalities. This model is supposed to take advantages of both biomolecular and quantum paradigms and to overcome some of their inherent limitations. We extend a recently proposed formal model of interface between a me...

This work presents a hybrid model of high perfomance computations. The model is based on membrane system (P system) where some membranes may contain quantum device that is triggered by the data entering the membrane. This model is supposed to take advantages of both biomolecular and quantum paradigms and to overcome some of their inherent limitatio...

We illustrate a quantum-inspired solution to the following dictionary search problem: given as input a word from a predefined nat-ural language (for example, English), produce its translation to another predefined natural language (in our example, Romanian). In our solution, words are represented as vectors in an appropriate finite-dimensional Hilb...

In this work the trends of modern developments of the auto- matic annotations of medical images are presented and analyzed. The specific selection of discussed trends for creation and usage of medical image annotation in SonaRez system is presented.

During the last decade both the medical images collection hardware and processing software have gotten the advanced development. As the result, medical imaging meets the challenge of treating extremely large medical images databases as well as complex algorithms. Such challenges are the natural tasks for massive parallel processing that can be prov...

Two main approaches are used in the development of computer-aided decision support systems in medicine: (i) based on the processing of images obtained during medical investigations, (ii) based on rules. At present, various combinations of these two techniques become more adequate and assure a great perspective for this domain. SonaRes system combin...

Despite the difficulties of ultrasound image interpretation, this source of information is still significant for diagnosis decision making. This paper describes the experience of SonaRes diagnostic decision support system (DDSS) for ultrasound examination development. Considering two-layer structure of the information contained in ultrasound images...

We describe the investigations on natural computing in the Institute of Mathematics and Computer Science of the Academy of Sciences of Moldova during last fifteen years. Most of these investigations are inspired by results and ideas belonging to Corresponding Member of the Romanian Academy Gheorghe Paun.

Ultrasound image is the primary (input) information for every ultrasonic examination. Since being used in ultrasound images analysis the both knowledge-base decision support and content-based image retrieval techniques have their own restrictions, the combination of these techniques looks promissory for covering the restrictions of one by advances...

The work describes approaches to development of tools for the Web portal on musical arts in Moldova. The tools are based on open sources, contain a simple intuitive interface to administrate the portal, and permit execution by a specialist in the area of musical arts of the content management operations: addition, modification, deleting, etc., over...

Membrane computing is a formal framework of distributed parallel multiset processing. Due to massive parallelism and exponential space some intractable computational problems can be solved by P systems with active membranes in a polynomial number of steps. In this paper we generalize this approach from decisional problems to the computational ones,...

Evolution of the informatization notion (which assumes automation of majority of human activities applying computers, computer networks, information technologies) towards the notion of {\it Global Information Society} (GIS) challenges the determination of new paradigms of society: automation and intellectualization of production, new level of educa...

A decision support system SonaRes destined to guide and help the ultrasound operators is proposed. The system is based on rules and images and can be used as a second opinion in the process of ultrasound examination.

Experimental digit form information about global ionosphere state is transmitted from the satellite by telemetric channel, is accepted by earth point of receiving and is processed by computer tools complex. Authors provided the investigation of information flows in the system from standpoint of total time of processing and possible information loss...

At the Institute of Applied Geophysics (Moscow) the system of information maintenance (SIM) of geophysical situation control is created. The creation of geliogeophysical informational-calculational system in limited time and with hard restrictions on expenses requires the preliminary quantitative evaluation of its building characteristics. In the p...

A decision support system SonaRes dedicated to guide and help the ultrasound operators is proposed and compared with the existing ones. The system is based on rules and images and can be used as a second opinion in the process of ul -trasound examination.