Comparative Analysis of COVID-19 Images through different Filtering Techniques using PSNR, MSE and SSIM

Lung diseases are a significant global health concern, and early detection is imperative due to their contagious nature. Integrating computer-aided detection technology with im-aging techniques can significantly enhance diagnostic accuracy. Preprocessing techniques are pivotal for refining image quality. Research has been conducted to identify the most suitable preprocessing filters for COVID-19 images obtained from hospitals. Evaluation techniques such as Peak Signal-To-Noise Ratio (PSNR), Structured Similarity Index Meth-od (SSIM), and Mean Square Error (MSE) are utilized to assess image quality. This study aims to determine the optimal techniques for enhancing image quality and improving diagnostic precision on private datasets. In conclusion, 62 images exhibited superior results with the bilateral filter, while the remaining 41 images demonstrated optimal outcomes with the Gaussian filter when analyzing images collected from the hospital.