Real time garbage detection using CNN and YOLO algorithms
S. Dhanushiya, K. Navaneetha, V. Vijayalakshmi · 2025
Modern metropolitan environments require real-time trash management since poor rubbish disposal can result in several environmental problems. In this paper, a novel garbage detection system based on convolutional neural network (CNN) and You Only Look Once version (YOLOv8) algorithms is proposed. By the use of live video feeds from webcams or CCTV, the system is intended to detect people handling rubbish and identify and categorize it. YOLOv8 is utilized to record video frames and identify things, and a CNN is used to categorize the objects into distinct garbage groups. A pre-trained dataset is used by the suggested method to attain high accuracy, with a mean average precision (mAP) of 0.94 and a 94% detection accuracy. According to experimental data, this method is effective at automating waste management operations, which may find use in smart cities.