Comparison of Neural Network and SVM classification with GLCM features extracted for Brain MRI Images

According to many Casualties due to a brain tumor, early detection for treatment and reduction of mortality in the early stages of are required. Due to the high complexity of brain tissue, manually detecting of brain tissues and tumor is very time consuming and depends on the operator. In this paper, we will use co-occurrence Matrix GLCM for texture detection and extraction of image feature. As well as to improve the accuracy of classification algorithms to determine the statistical characteristics and characteristics of the tissue from the brain tumor we used: contrast, correlation, energy and homogeneity. The results show that the method of image feature extraction (GLCM) is better performance than other methods and the use of these features in classification algorithms to improve accuracy, especially compared to the MLP and neural networks

Keywords:

Image Classification , Neural Networks , GLCM , SVM