Intelligent Clothing Matching Based on Feature Analysis

Clothing feature analysis based on traditional image content feature retrieval cannot express the deep semantic features of the image, so this paper proposes an intelligent method of clothing matching based on fusion feature analysis. Firstly, the clothing features extracted by the deep convolution neural network can fully express the whole and detail semantic attributes of clothing, and the metric learning is realized using trip let loss. Then, the database of clothing matching is generated by target detection to process the data set of clothing matching. Finally, by constructing a disordered matching network, the clothing combination is regarded as a clothing set, and the matching recommendation is realized by analyzing the clothing information distribution. The experimental results show that the proposed algorithm is better than the single algorithm, and it can recommend suitable matching schemes for users according to clothing image and matching algorithm.