Seyed Hashem Davarpanah - Assistant Professor Dep. of Computer Engineering Birjand University of Technology Birjand, Iran

This paper introduces an automated waste segregation instrument commercially productive for laboratory applications and for medical waste recognition. In this system, a fuzzy rule based system is designed being able to segregate 24 different hospitals waste types which are selected from 8 more common medical waste groups. After capturing each frame, some preprocessing operations are done, features are extracted, fuzzy parameters, fuzzy terms and fuzzy rules are determined and finally a rule with a maximum certification degree is fired. The system is flexible on different light conditions, view degrees of the camera, and 360 degree object rotation. Our experiments on a real environment and using both new and used wastes show that the performance of the system is more than 94%. It works on different light conditions, different view degree of camera and different positions of the objects.

component; medical waste segregation; fuzzy expert system; fuzzy rule-based system