Using Principal Component Analysis and Radial Basis Function Neural Networks for Image Contents Classification
Chuan‐Yu Chang, Chi-Fang Li, Hung‐Jen Wang · 2007
Due to the vogue of digital cameras, it is easy to obtain digital images. And with the rapid development of digital image processing, database and internet technologies, how to efficiently manage a large amount of digital images become very important. Therefore, in this paper, we propose a novel method, which integrates the principal component analysis (PCA) and modular radial basis function (MRBF) neural networks for semantic-based image content classification. Since the traditional RBF neural networks are sensitive to center initialization and time- consuming in the training procedure. To overcome these drawbacks, a combination of self-organizing map (SOM) neural network and learning vector quantization (LVQ) is used to select more appropriate centers for RBF networks, and a specific designed MRBF neural networks is used to improve the classification accuracy and accelerate the training time. Experimental results show that the proposed method is capable of analyzes the component of photograph into semantic categories with high accuracy and the result of photograph analysis is similar to human perception.