Medical Waste Image Classification System Using Deep Learning

Sorting medical waste is a crucial phase in making sure that healthcare facilities take care of and dispose of garbage responsibly. The spread of infection poses a significant risk that leads to a public health concern in the community. This research aims to create an image classification automated system using deep learning model like InceptionV3, which can accurately predict whether a given medical waste image is infectious or not. Data curation was use to obtain 534 images which was selected, organized, and gathered from TrashBox dataset available on GitHub. InceptionV3 with no feature extraction has the highest accuracy (0.91), precision (0.89), recall (0.91) in the identification of medical infectious waste. The evaluation can have a potential in contributing to the proper sorting and disposal of medical waste by hospitals, promoting a cleaner environment. 

Keywords: medical waste, deep learning, neural network 

Corresponding author's email: srcb.laureterollymcjames@gmail.com