Abstract: With the advancement in web technology, search engines are exploited daily by users in search for all kinds of information. However, most search engines are text format and do not provide adequate searches for images. In this paper, we propose a scheme to an image oriented search engine. Images will undergo discrete cosine transformation (DCT) from the spatial domain to the frequency domain. The low and middle frequency domains will then be used to construct two levels of eigenvalue comparisons. Test results showed that users could effectively search images from the image database using our scheme.
Jeanne Chen , Tung-Shou Chen , Kwang-Fu Li and Zhi-Haur Wu , 2004. An Eigenvalue Comparison Search Engine for the Discrete Cosine Transformed Image . Asian Journal of Information Technology, 3: 589-598.