Alcione Jandir Candeas,  Ulisses de Mendonça Braga Neto,  Edson Costa de Barros Carvalho Filho. 

We deal in this work with a broadly studied topic in astronomical imaging, namely the star/galaxy discrimination problem. Through the use of Mathematical Morphology, we propose an original approach to the characterization of these astronomical objects, which is based on gray-level shape information. The main steps in our method are image pre-processing, segmentation and feature extraction, all of which employ Mathematical Morphology tools that were implemented using the MMach toolbox for the Khoros system. We present a comparison between our segmentation results, based on the watershed method, and those of a classical software package SExtractor. The shape information is extracted through the use of the gray-level morphological pattern spectrum, which yields very satisfactory shape features, that promise to be very suitable for future work in neural-network automatic classification.

http://sibgrapi.sid.inpe.br/rep-/dpi.inpe.br/ambro/1998/05.06.14.10

Processando consulta.