The improper management of bio-medical waste generated by healthcare facilities and research facilities harms the environment and the public’s health. The daily agenda used by the traditional trash management method is quite ineffective and expensive. Due to improper recycling methods, the public has also demonstrated the inefficiency of the current recycling container. The existing human sorting process is inefficient and dangerous for waste disposal and garbage carter, requiring the development of a system to identify and classify bio-medical waste materials. The recognized issue in the current system emphasizes the difficulty and risks of manual sorting. To increase the safety, efficiency, and sustainability of biomedical waste management procedures, the current study is based on deep learning that uses the perceptron model to precisely detect and classify objects that are part of biomedical waste. Deep learning’s fundamental idea is that digital systems need to function similarly to how humans do. The suggested method might eventually accomplish an overall 90% accuracy rate, leading to expense savings and reduced risks associated with manual sorting.