Search

This work proposes a methodology of digital analysis of the state of conservation of a canvas painting to solve difficulties related to the pictorial reintegration of paintings that present an excessive number of different lacunae in terms of size and extent. The case study is related to the small-size oil on canvas painting executed by an unknown artist, where the lacunae were challenging to analyze and localize graphically. Therefore, it required a careful evaluation of the approach to be used during the pictorial reintegration intervention. Using an image analysis method, based on the semi-automatic extraction approach, the state of conservation’s graphic relief outlining different virtual operating proposals was obtained.

Questo lavoro propone una metodologia di analisi statistica delle immagini diagnostiche finalizzata a migliorarne la lettura e a facilitare la trascrizione grafica dello stato di conservazione di beni artistici, rendendola puntuale e ripetibile. Si presenta come caso di studio un piccolo dipinto ad olio su tela di autore ignoto in cattivo stato di conservazione. Utilizzando il metodo citato, basato su un approccio semi-automatico di estrazione delle aree di interesse, si otterranno delle schede di rilievo relative allo stato di conservazione con le quali sarà possibile eseguire statistiche zonali per calcolare la percentuale dell'area danneggiata rispetto all'intera superficie del dipinto. Le operazioni mostrate possono essere applicate ad ogni tipologia di immagine diagnostica, studiando in maniera pi? oggettiva lo stato di conservazione di qualsivoglia manufatto.

The accuracy of the naval map representation determines the quality of route construction when solving the problem of automatic control of ships movement in a difficult navigation environment. Using a high-precision raster map requires a large amount of memory to store data. Vector representation of the map is an alternative approach to the problem of storing data about terrain, it allows to store only the data containing the terrain contours. In this paper we implement and research the algorithms which allow to vectorize the raster maps by the set of isoline slices using irregular grid of heights. These algorithms approximate the map with a set of elevation slices, recursively search for nested isolines on the slice, build and optimize contours of the isolines on the slice. We conduct a study that shows a significant reduction in the amount of stored data in vector form compared to raster format. The proposed format for storing map data will be used to represent realistic landscapes in the tasks of building ship routes.

The problem of raster images vectorization was analyzed and researched in this work. Existing vectorization systems are quite expensive, the results are inaccurate, and the manual vectorization of a large number of drafts is impossible. That‘s why our goal was to design and develop a new raster images vectorization system using our suggested automatic vectorization algorithm and the way to record results in a new universal vectorial file format. The work consists of these main parts: analysis, design, user documentation and experimental research. Conclusions of the analysis of this field are that the optimal solution of vectorization is the synthesis of automatic and manual vectorization. Users’ requirements were documented and raster image vectorization system designed. It was noticed that the result of vectorization depends on: vectorization algorithm; vectorization parameters; quality of raster image and user experience in this field. The core problems of vectorization are: symbol, intersection and corners recognition, lines‘ and their endings‘ types, filled areas. Vectorization results of our system and Corel TRACE were compared during the experimental research.