关键词: 材质识别, 深度卷积神经网络, 智能衣物洗护

Abstract: With the growth of the high-net-worth population, traditional clothing care methods are no longer able to meet people's pursuit of quality. Intelligent clothing care technology can intelligently match care programs by accurately identifying clothing materials, satisfying the demand for high-quality care. This article establishes a clothing material recognition method based on deep convolutional neural network image recognition. It first recognizes the type of clothing and then further identifies the material of the clothing, achieving fast and accurate material recognition and clothing classification. Testing was conducted on over 4000 images of clothing taken in a clothing care cabinet according to the test script. The results show an accuracy of 92.06% in clothing type classification, 94.00% in material classification, 87.27% in identification accuracy, and 100.00% accuracy in cases with no clothing identified. This clothing material recognition method exhibits high accuracy and strong generalization ability, can providing technical support for intelligent clothing care programs, possessing broad application prospects.

Key words: Material recognition, Deep convolutional neural network, Intelligent clothing care