A Study On Emotional Automatic Classification Of Fabrics In Website Of Electronic Commerce

Emotional semantic classification of images is a very important field in semantic-based image retrieval. Emotional semantics, as an important high-level semantics, occupies an important position in image retrieval based on semantics. It is very necessary to classify images according to their emotions. Traditional image classification method mainly classifies images according to the similarity of visual features, and often ignores the influence and effect of image emotion. It cannot meet the real needs of users. The classification of image according to emotion semantics will greatly improve the image classification based on Semantic image retrieval performance.Different clothing having different emotion, fabrics also have their own emotions, clothing designers need to consider the feelings expressed by the fabric when they design, and consumers in the purchase of clothing will be based on t heir emotional needs to buy goods. By the rapid development of E-commerce in recent years, more and more online shopping platform coming to people's life. We can buy almost anything from anywhere. This greatly reducing the distance between consumers and businesses. People can easily select goods through the network, but it is difficult to select what you want from a large number of information. We need to carry out a reasonable classification of goods.This paper studies the automatic identification of fabri cs emotion in e-commerce sites, and puts forward a method of automatic recognition fabrics commodity emotional achieving fabrics commodity classification and sorting according to emotion. In the previous study, the emotion of the fabric has been divided into seven sensible words with opposite semantic meaning. Through the extraction of the underlying features such as the color and texture of the fabric images, and using the support vector machine tech nology to learn the emotion of the images,we can get the emotion factors, then calculates the quantitative description of the emotion. The main research content of this paper is to automatically recognize the emotion of the fabric image, and apply this recognition technology to the website of electronic business, so that people can choose the fabrics according to their emotions conveniently. Through the study of image feature extraction and classification algorithm, this paper develops an electronic business website which can automatically identify the emotion of fabric, and achieves good results by experiments. It proves the practical value of this paper. It is proved that the study of emotion classification of fabric images is feasible. This topic has important theoretical and practical value.