Brain Tumors Detection on MRI Images through Extracting HOG Features
DOI:
https://doi.org/10.46947/joaasr21201896Keywords:
Tumor detection, Histograms of Oriented Gradients, Feature extraction, Support Vector MachineAbstract
In this paper, a system of Computer Aided Diagnosis (CAD) is introduced for detecting brain tumors through MRI images. The proposed system consists of three main sections: segmentation, feature extraction, and categorization. In the segmentation phase, a system called Seeded Region Growing (SRG) is applied to separate brain tissue from other regions. In the feature extraction phase, the Histograms of Oriented Gradients (HOG) algorithm is used. Finally, in order to describe the images of the brain, we use Support Vector Machine (SVM) and classify the images in two tumor-free and tumor-grade groups, and then compare them with similar tasks. The results show a high efficiency of this approach compared to other methods. 850 MRI images have been applied to test and teach samples; the number of healthy brains is 300 and the number of defective brains is 550, and the system has achieved a precision rate of 93.2% in the categorization.
Metrics
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.