Att ämnesbestämma bilder
En teoretisk undersökning
The aim of this master ´s thesis is to understand why there are difficulties in finding joint methods and standards in the field of image indexing. The main focus for interpreting these issues is on Sara Shatford ´s article Analyzing the Subject of a Picture: A Theoretical Approach. In the article Shatford discusses a theoretical basis for identifying and classifying the subject in a picture. In doing so she uses concepts taken from The Philosophy of Art, Meaning in Language and Visual Perception. With the aid of S.R. Ranganathan ´s Colon Classification she models a figure which is meant to be used as a tool when analysing the subject of a picture. The main focus is on the concept-based information retrieval field that sets its aim on the intellectual/textual methods for indexing images. By way of introduction I examine the concept of a subject, thereafter this method for subject analysis of a picture is being explored. To make the focal point broader The Art and Architecture Thesaurus is examined and thereafter compared to Shatford ´s theories in order to lay them open for criticism. The conclusion from this examination is that Shatford ´s model may be used as a tool for the subject analysis of a picture. I do nevertheless find some of Shatford ´s approaches problematic. One is her applying Panofsky ´s method for analysing a picture in which it is not possible to analyse the specific purpose of a picture. In my opinion Roland Barthes´ semiotic methods for analysing a picture are more accurate. Another question that occurs when indexing images is whether it is desirable to index with the aid of language. Automatic tools such as content-based information retrieval CBIR might be the answer to this.