作者: Pınar Çivicioğlu , Mustafa Alçı , Erkan Beṣdok
DOI: 10.1007/978-3-540-30198-1_39
关键词:
摘要: In this paper, a new filter, RM, which is based on exact radial basis function artificial neural networks, proposed for the impulsive noise suppression from highly distorted images. The RM uses Chi-Squared goodness-of-fit test in order to find corrupted pixels more accurately.The filter shows high performance at restoration of images by noise. extensive simulation results show that achieves superior other filters mentioned paper cases being effective and detail preservation, especially when density very high.