A novel recycling method using machine vision to assist in the processing of stacked waste fans

With the increasing amount of Waste Electrical and Electronic Equipment (WEEE), waste electric fans are an important component that cannot be ignored due to their variety and high recycling value. Utilizing intelligent methods for waste recycling has become a prominent topic in contemporary society. However, there is a lack of sophisticated datasets and algorithms in the field of waste electric fan recycling. In order to help automatically recycle a large number of piled-up electric fans, this paper proposes a novel Mrcnn-Stafans instance segmentation model based on a modified mask R-CNN. In addition, a stacked electric fan-TRASH dataset that is closer to the actual recovery situation is constructed. Experimental results show that the proposed improved algorithm achieves 98.161% accuracy while optimizing the number of model parameters. The focus of this work is to help identify recycled WEEE fans using visual recognition techniques, thus helping solve environmental problems.