Ekstraksi Ciri dan Identifikasi Citra Otak MRI Berbasis Eigenbrain Image

Autor: Soesanti, Indah, Susanto, Adhi, Sri Widodo, Thomas, Tjokronagoro, Maesadji
Jazyk: angličtina
Rok vydání: 2011
Zdroj: Forum Teknik; Vol 34, No 1 (2011)
ISSN: 0216-7565
Popis: In this research, we exctract and identify MRI brain images based on eigbrain image. MRI barain images are used to be input for feature exctraction and identitication. Feature exctraction is done by using the eigbrain image. For all reference image, we find image mean and eigbrain image, and the results are stored. If there is test image, we will find the nearest distance of eigenbrain between test image and reference images. The feature extraction is used to identify the image is whether the normal brain image or the brain image with tumor.The results show that the method successfully classifies MRI images into tree clusters: normal, glioma, and metastase. The input test images can be identified accurately 100% for image sizes from 256 x 256 pixels to 64 x 64 pixels.Keywords : feature extraction, image identification, MRI medical image, eigenbrain image.
Databáze: OpenAIRE