Efficient Automated Detection System for Pneumonia-Affected X-Ray Images using Neural Networks

COVID-19, caused by the SARS-CoV-2 virus, has impacted over 250 countries worldwide. Early recognition of COVID-19 infections is essential to effectively manage this global outbreak. While chest X-rays are widely available and cost-effective imaging techniques, their accurate interpretation necessitates the expertise of knowledgeable professionals in assessing the severity of infectious. Even though many COVID-19 pneumonia virus recognition techniques are available, each method has advantages and disadvantages. This paper proposes an efficient automated detection system for pneumonia-affected X-ray images utilising a bilateral filter and RF classifier. The proposed method presented in this study was trained and evaluated using a publicly available X-ray dataset specifically designed for classifying tertiary and normal cases. Furthermore, the proposed method demonstrated reliable accuracy of 94% and recall of 87% in binary classification tasks.