期刊名称：International Journal of Computer Science Issues
印刷版ISSN：1694-0784
电子版ISSN：1694-0814
出版年度：2014
卷号：11
期号：3
出版社：IJCSI Press
摘要：Brain tumor has ample variations from person to person, in terms of size, texture or distribution. Experienced oncologists can easily identify the tumor region. But given its variations, training an algorithm to detect the tumor is a challenging task. Many algorithms have been developed to segment and detect the tumors, taking into consideration various aspects of an image. In this paper, we discussed a survey of brain tumor segmentation algorithms in image processing, and our proposed approach to detect brain tumors using texture and feature vectors and one-class classification methods.
关键词：Brain Tumor; Image Segmentation; Histogram of Oriented Gradients; One Class Classification.
