Back to Volume
Paper: Transforming Images into Icons to Remotely Retrieve Information from Astronomical Archives
Volume: 101, Astronomical Data Analysis Software and Systems V
Page: 509
Authors: Csillaghy, Andre
Abstract: As astronomical image archives increase in size, information is getting more intricate to access. To compensate for this drawback, software methods to query the archive by its image content must be developed. Innovations in content-based information retrieval systems will eventually speed up and simplify the process of finding relevant information. To this end, the method presented here extracts features by considering data density in a three-dimensional attribute space. Relevant features are stored in symbolic representations called icons. Icons not only make possible queries by browsing or comparing, but also give the possibility to efficiently transfer information contained in images through the network.
Back to Volume