One of the most popular medical diagnostic tools ever is X-ray imaging, and that's saying something considering how far biomedicine has come.The amount of information that may be retrieved from images is significantly impacted by the characteristics that are already there.We propose to apply the Euclidean distance transform technique for image preprocessing to take the advantage of the image feature technology that makes it easier to identify pneumonia cases.The paper concurrently applya number of image preprocessing techniques, such as binarization, thresholding, scaling, normalization, and others, to the sampled image before the features are obtained. Then we extract the characteristics for image classification and recognition.The suggestion is verified and examined using the publicly accessible COVID-19, Pneumonia, and Normal image datasets.