Automatic source detection in astronomical images

Abstract

The use of automatic algorithms to detect astronomical sources (stars, galaxies, gas, dust or cosmic rays) becomes of great importance because of the increasing amount of astronomical images and the inefficiency and inaccuracy of manual inspection. We exhaustively analyze the state of the art on this topic, presenting a new classification of techniques and pointing out their main strengths and weaknesses. A complementary quantitative evaluation of some of the most remarkable methods found in the literature is also provided. Moreover, we present three different proposals based on different strategies to detect faint sources in radio images, and another proposal to deal with infrared and radio images (characterized by having a complex background). The experimental results and the evaluation performed with synthetic and real data points out that our new proposals perform better than state­of­the­art approaches in terms of reliability and completeness of the detections provided.