Pneumonia Classification using SVM and GLCM, LBP and HOG Features

A type of lung disease is Pneumonia that infect the respiratory system, and this disease can be recognized by X-ray images for the chest area. The presence of pneumonia or not need the medical professional to examine the X-ray image, and the rule of his decision not ignored, but may suffer from some problem that can led to mistake. Many studies introduced in this field by using new technology, artificial intelligent and deep-learning technics to reduce the error ratio. In this work we introduce a model that  correctly classify the chest X-ray as a pneumonia or not,  GLCM, LBP and HOG are used as a feature extraction methods after preprocess the used images. SVM used here as a classifier and the obtained results show the effectiveness and robustness SVM when used with a suitable preprocessing steps and prices  feature extraction methods. The model Accuracy is 99.90% with Recall of 100%, Precision : 99.87 %, and F1-score: 99.94%. The used dataset is chest X-ray from  kaggle with 5,216 chest X-ray images