Image Emotional Classification Based on Color Semantic Description

摘要

Describing images in semantic terms is an important and challenging problem in content-based image retrieval. According to the strong relationship between colors and human emotions, an emotional image classification model based on color semantic terms is proposed in this paper. First, combined with PSO, fuzzy c-means clustering is implemented for color segmentation, and eight color clusters can be obtained to describe the main color of an image. Secondly, based on Wundt's theory, a 3D emotional model is constructed and a novel approach for describing image color semantic is proposed. Finally, we present a trial classification system which allows users to query images using emotional semantic words. Experimental results demonstrate that this model is effective for sentimental image classification.