作者: Mahmoud I Khalil , Mohamed M. Bayoumi , None
DOI: 10.1109/34.954605
关键词: Artificial intelligence 、 Mathematics 、 Invariant polynomial 、 Wavelet transform 、 Conic section 、 Wavelet 、 Invariant (mathematics) 、 Applied mathematics 、 Pattern recognition 、 Affine transformation 、 Invariant measure 、 Edge detection
摘要: Dyadic wavelet transform has been used to derive an affine invariant function. First, function using two dyadic levels is derived. Then, this another six levels. We introduce the based conic equation. The on analyzing object boundary transform. Experimental results both synthetic and real data are demonstrate discriminating power of proposed It also compared with some traditional methods. stability examined. In addition, under large perspective transformation tested.