Improving Retrieval Performance of Zernike Moment Descriptor on Affined Shapes

Dengsheng Zhang and Guojun Lu
Gippsland School of Computing and Information Technology
Monash University
Churchill, Victoria 3842
Australia

Tel: 61-3-9902 6772
Fax: 61-3-9902 6842
Email: dengsheng.zhang, guojun.lu@infotech.monash.edu.au

ABSTRACT

Shape (object) description consists a key part of image content description in MPEG-7. In the development of MPEG-7, Zernike moment descriptors (ZMD) have been adopted as region-based shape descriptors. However, because Zernike moments are defined in a unit disk and extracted from shape in polar space, less shape information is made used for skewed and stretched shapes. As a result, the extracted ZMD is not robust to skew and stretching distortions. In this paper, an enhanced Zernike moment descriptor (EZMD) is proposed to overcome this drawback. The EZMD is acquired by deriving ZMD from the rotation and scale normalized shape. The rotation and scale normalization effectively suppresses the uneven shape distribution within the unit disk and compensates the skew and stretching. Consequently, more shape information is made used in the feature extraction. Experimental results show that the proposed EZMD outperforms ZMD significantly. (full paper)

Key words:Zernike moments, shape, CBIR, retrieval.