A rotation-invariance feature of an amplitude-coupled minimum-average correlation-energy (MACE) is investigated and compared with that using a simple MACE algorithm. The amplitude-coupled MACE filter is studied in terms of the statistical probability of correct classification between the in-class and out-of-class SAR targets. The amplitude-coupled MACE algorithm offers a much better rotation invariance than a simple MACE algorithm. The rate of probability of correct classification is also very high. (AIAA)
- Rotation-invariant target recognition using an amplitude-coupled minimum-average correlation-energy filter
- Khan Iftekharuddin - Dominion (United States)Farid Ahmed - University of DaytonMohammad Karim - University of Dayton
- Optical engineering, Vol.35(4), pp.1009-1014
- 6
- Department of Electrical and Computer Engineering
- English
- Journal article
- https://doi.org/10.1117/1.600717
- 9914541779001301