Weighted aspect moment invariant in pattern recognition
Many drawbacks has been found in Hu's moment Invariant or known as Geometric Moment Invariant (GMI). Due to its flexibility, GMI is still widely used by the researchers until now. This paper proposes an alternative approach, Weighted Aspect Moment Invariant (WAMI) by combining Weighted Central...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Book Section |
| Published: |
Springer Verlag
2009
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/13193/ http://dx.doi.org/10.1007/978-3-642-02457-3_66 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Many drawbacks has been found in Hu's moment Invariant or known as Geometric Moment Invariant (GMI). Due to its flexibility, GMI is still widely used by the researchers until now. This paper proposes an alternative approach, Weighted Aspect Moment Invariant (WAMI) by combining Weighted Central Moment (WCM) and Aspect Moment Invariant (AsMI) to solve GMI's drawbacks in term of noise and unequal data scaling. Various insect images are used in this study with two different sizes as simulation images. The simulation results show that the proposed WAMI improves inter-class and intra-class criteria for unequally scaling data compared to AsMI. |
|---|