A content-based image retrieval system using RBF neural network

作者：Lu Yinghua^*; Tang Changhua; Kong Jun; Zhao Qiushi; Li Bingbing

来源：1st International Conference Bio-Inspired Computing -Theory and Applications, 2006-09-18 to 2006-09-22.

• content-based image retrieval
• region-based
• RBF
• PCA

This paper presents a novel image retrieval system using radial basis RBF) neural network that splits the retrieval process into two stages. In Stage I of the system, we use Global Color Histogram (GCH) to represent the color infomation, and wavelets and Gabor are used, to represent the texture information. The combined global color and texture features am extracted, and consequently, the dimensionality of feature vectors is reduced by principal component analysis (PCA). In order to reduce searching time, we categorize various patterns using RBF network. In the second stage, the region-based image presentation is employed and for each corresponding category or group, the relevant images are retrieved. Finally, we compare the performance of the proposed system with other systems. Experimental results show the efficacy of the proposed system.

• 出版日期2007-8
• 单位东北师范大学