Pneumonia Detection using X-ray Image Analysis with Image Processing Techniques

Abstract

Chest X-ray (CXR) examination is the frequently used diagnostic image technique to detect pneumonia infection. Examination of CXR is a challenging task even for trained radiologists due to the presence of noise, contrast issues, and subjective variability. This paper focuses on a computer-aided diagnostic method for Pneumonia using CXR images that can enhance the efficiency of diagnosis. Many existing image segmentation and edge detection techniques have been tried to segment the lung; however, none of the existing methods generated a promising segmentation output. This paper proposes an efficient method of identifying the presence of pneumonia mists in chest X-rays by an integrated approach of Gaussian filter, CLAHE contrast improving technique, and Sobel edge detection. A comparative study of the entropy values for all the edge detector operators were carried out and the results were validated with our proposed model. The Sobel operator is found to have better results compared to other traditional methods.