摘要： 提出了利用融合不同的低层MPEG-7视觉描述符的方法来进行基于内容的图像分类的技术。目的在于通过融合几种描述符来改善机器学习分类器的性能，包括3种方法来改善分类器的性能：作用于支持矢量机（SVM）分类器的聚类融合，作用于K近邻分类器的反向传播（BP）融合和作用于FART模糊神经网络的BP融合。将这些分类方法应用到海滩风景/城市风景的分类的实验中，实验结果表明BP融合显示出更好的性能改善。

关键词: 视觉描述符, 分类器, 融合, 特征提取, 模糊神经网络

Abstract: Several content-based image clas¬sification techniques based on fusing various low-level MPEG-7 visual descriptors are proposed in this paper. The goal is to fuse several descriptors in order to improve the performance of several machine-learning classifiers. Three approaches are described: A “merging” fusion combined with an SVM classifier, a back-propagation fusion combined with a K-Nearest Neighbor classifier and a Fuzzy-ART neurofuzzy network. In the latter case, fuzzy rules can be extracted in an effort to bridge the “semantic gap” between the low-level descriptors and the high-level semantics of an image. Experimental results on the beach/urban scenes classification problem show the best.

Key words: Visual descriptor, Classifier, Fusing, Feature extraction, Neurofuzzy network