Tehnički vjesnik, Vol. 15 No. 4, 2008.
Prethodno priopćenje
Segmentation of the CT image using self-organizing neural networks
Marko Martinović
; Veleučilište u Slavonskom Brodu, Trg S. Miletića 12, 35000 Slavonski Brod
Antun Stoić
; Veleučilište u Slavonskom Brodu, Trg S. Miletića 12, 35000 Slavonski Brod
Darko Kiš
; Poljoprivredni fakultet u Osijeku, Trg Sv. Trojstva 3, 31000 Osijek
Sažetak
A presentation of the segmentation and extraction process of CT (Computed Tomography) images, using self-organizing neural networks has been given. The image features have been used during segmentation process and adopted for each pixel on image. The origin of the features, used in this paper, are the scaled differential-geometrical features which are the entering data for neural network. Neural network, after training, dissipates the pixels into one of the image regions.
Ključne riječi
CT image; features; neural networks; segmentation
Hrčak ID:
31409
URI
Datum izdavanja:
31.12.2008.
Posjeta: 1.898 *