Blur invariants: A novel representation in the wavelet domain |
| |
Authors: | Iman Makaremi Majid Ahmadi[Author vitae] |
| |
Affiliation: | Department of Electrical and Computer Engineering, University of Windsor, Windsor, ON, Canada N9B 3P4 |
| |
Abstract: | Blur invariants in the wavelet domain are proposed for the first time in this paper. Wavelet domain blur invariants take advantage of several benefits that this domain provides, i.e. different alternatives for wavelet function and analysis in different scales. It is not required to model the blur system in order to extract the invariants. It will be shown how the space domain blur invariants are a special case of the proposed invariants. It will also be explained how the proposed invariants would not have the null space that their special case in the spatial domain have which limits their discriminative power. The performance of these invariants will be demonstrated through experiments, and compared to its counterpart which is defined in the spatial domain. |
| |
Keywords: | Blur invariant moment Direct analysis Feature extraction Wavelet transform |
本文献已被 ScienceDirect 等数据库收录! |