Methods for astronomical image restoration

Ana Paula Abranches de Castro e Shiguemori¹; Elcio H. Shiguemori²; Murilo da Silva Dantas¹; Alice Kozakewicius³; Vinicius Monego³; Renata Sampaio da Rocha Ruiz⁴; Cristiano Strieder⁴; Haroldo Fraga de Campos Velho⁴

¹ IFECT Sao Jose dos Campos; ² DCTA; ³ UFSM; ⁴ INPE

doi:10.20906/CPS/CILAMCE2017-1229

Resumo

Image restoration is a kind of inverse problem of great relevance. It is one of the most important topic in image processing on space applications. Remote sensing and autonomous navigation by images are typical applications. Another important application is in astronomy. The focus of this paper is the evaluation of different restoration techniques based on truncated singular value decomposition, Tikhonov regularization method, neural network restoration, and wavelet filtering. The image restoration techniques are applied in astronomical images of planets of the solar system and from solar activity.

Palavras-chave: Image restoration; Astronomical images; Inverse problem