Design and Implementation of Biomimetic Intelligent Waste Sorting Bin

Waste classification is essential for establishing a good living environment and maximizing resource utilization. However, the actual effectiveness of waste sorting remains suboptimal due to the public's lack of knowledge and weak awareness of classification. To address these, this work develops a bionic intelligent waste-sorting bin based on the YOLOv5 neural network classification algorithm. Specifically, the overall structure of the bin is first designed and key hardware components are selected. Next, this work builds the mechanical structure and hardware circuits of the bin. Using the YOLOv5s algorithm, this work further improves the adaptive image scaling algorithm to enhance the accuracy of waste classification. Finally, this work designs an app using frontend technologies such as HTML, CSS, and JavaScript to monitor the bin in real time. Through extensive testing, this intelligent waste-sorting bin accurately classifies and collects four types of waste, monitors the bin's overflow status in real time, and promptly notifies staff, greatly reducing the labor involved in the sorting process. This solution offers practical value for waste classification, recycling, and conserving societal resources.