The Study Of Clothing Image Emotion Classification Based On Multi-Features

Image emotion semantic classification is an important and challenging task in the field of semantic-based image retrieval. The image contains abundant emotions, it is necessary to classify these images according to emotions which expressed. The traditional technology of image classification mostly retrieves images by analyzing the similarity of image visual features, however, it neglects the impact and function of image emotion, which can not meet users' real needs, so classifying images rationally using image emotion semantic will greatly improve the performance of semantic-based image retrieval systems.With the rapid development of material civilization and spiritual civilization, clothing has become an indispensable part of daily life, the purpose of costume designing is not only to prevent cold and hide one's shame, but also to meet and adapt people's psychological needs for clothing. So, emotion-based clothing image classification is imperative. In this paper, clothing images are data source, some important technologies and algorithms of clothing image emotion semantic classification system are studied deeply. "Clothing image semantic classification system" is designed and developed, which implements the clothing image classifying based on emotion semantic, we also make some meaningful discussions on how to choose clothing by its emotions.In this paper, image semantic model is introduced firstly, which is the abstraction of image's entire semantic representation and process procedure, and the theoretical basis for the topic is also provided. Secondly, the extraction methods of visual image features such as color, texture and shape etc are expounded. When features of clothing images are being extracted, integrating color and shape of clothing can express emotions commendably, so, in this paper, a method that fusing color and shape features is proposed, named planar histogram based color-edge direction angle fusion. Through societal investigating and document reading, we establish the relationship between low-level features and high-level semantic of clothing images, that is, the relationship between clothing's color-shape synergetic features and its style. That lays the foundation for the achievement of clothing image emotional semantic classification.This paper chooses (Radial Basis Probabilistic Neural Network algorithm, RBPNN) as classifying algorithm, because the new model is more advantageous than others with low computation complexity and fast convergent speed. This model can also be used widely in some fields such as pattern recognition and nonlinear function approximation and so on. At the same time, to the lack of this model, we propose a improved method that solves absolutely or partly the limitations of the neural network, and achieve a classification algorithm that can choose the best one from several activation functions.Through researching of image feature extraction and classification algorithm, a system of clothing image classification based on emotion is exploited, according to clothing images' color and shape feature values; we input these values into the inputting layer of neural network classifier, train and test classification of the neural network, which achieves a better result. The system's practicality is also validated.The whole paper' works proved that it is feasible to classify based on image's emotion, and it can greatly improve the accuracy of image classification using multi-feature integration and improved classification algorithm. The issue has important theoretical and practical significance.