Open AccessDomestic garbage recognition and detection based on Faster R-CNN
Author(s) -
Zhifeng Nie,
Wenjie Duan,
Xiangdong Li
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1738/1/012089
Subject(s) - garbage , sorting , computer science , identification (biology) , artificial intelligence , pattern recognition (psychology) , garbage collection , computer vision , algorithm , botany , biology , programming language
The core of intelligent garbage sorting is target identification and detection. In order to achieve effective garbage sorting, on the basis of deep learning, the Faster R-CNN target detection model and ResNet50 image classification model are used to identify and train 3984 garbage images, and predict 3552 images. The results show that the accuracy of garbage recognition is 89.681%, the average accuracy of each garbage prediction is 91.68%, and the accuracy of each category of garbage image prediction is over 93.3%. Through the identification, detection and classification prediction of garbage images, it provides data support for the intelligent classification of domestic garbage.