The Dataset was taken from kaggle.The original dataset was having high dimensional images and due to limited resource avilabilty,I have taken the cropped images which are of size 256 pixels of the original data.
The original images were in dicom format(;ie Digital Imaging and Communications in Medicine) and the file dicom_to_jpg.py file was used to convert the images.
Detailed explaination of each and every code is written in the notebook.I have integrated the weights and bias for experiment tracking and analyzing.
If there is opacity(;ie covid) then bounding boxes are plotted and if there is no chance of covid nothing is plotted
