A comparative study of Zernike moments

Abstract

Effective image retrieval by content requires that visual image properties are used instead of textual labels to properly index pictorial data. Shape is one of the primary low-level image features. Many shape representations had been proposed. The Zernike moment descriptor is the most suitable for shape similar-based retrieval in terms of computation complexity, compact representation, robustness, and retrieval performance. We study the first 36 Zernike moments and find the dependence relations between them. A new compact representation is proposed to replace the old one. It is not only saving storage capacity but also reducing the execution time of index generation.