Classification of Pneumonia using a Combined Approach of Image Processing and Deep Learning Algorithms

Abstract

The process of classifying and identifying pixel groupings or vectors in an image according to particular rules is referred to as image classification. With the expeditious development of science and technology and people's higher and better demand for quality of life, image automatic classification technology has been applied to numerous fields of development. When we classify the image, the normal image classification technique cannot accurately grasp the inner relationship between the identified objects, in addition to that because of the too high characteristic dimension of the data the traditional method also has the limitation of the recognition object feature expression. Also, it is necessary to choose which features are essential in each given image. Therefore the experimental results were not ideal. Considering the above mentioned content, this paper proposes an image classification technique gleaned from the convolutional neural network. The appropriate Convolutional Neural Network Model which is also called CNN is chosen for the dataset. Bacterial Pneumonia has comparable signs and symptoms to viral pneumonia. The deep learning methods/models that are proposed, distinguish Bacterial pneumonia to viral pneumonia. The images were pre-processed and trained for countless categorizations like Normal, Bacterial Pneumonia, and Viral pneumonia. The models advanced as a part of this work achieved about 98% accuracy on a test dataset which consists of three hundred images.