作者: Moncef Gabbouj , Faouzi Alaya Cheikh , Azhar Quddus
DOI:
关键词: Pattern recognition 、 Search engine indexing 、 Cascading Style Sheets 、 Mathematics 、 Artificial intelligence 、 Wavelet transform 、 Cognitive neuroscience of visual object recognition 、 Feature extraction 、 Invariant (mathematics) 、 Computer vision 、 Feature vector 、 Similarity measure
摘要: In this paper we propose a new approach for shape recognition using the wavelet transform modulus maxima. And apply it to problem of content-based indexing and retrieval fish contours. The description scheme similarity measure proposed here are simple take into consideration way our visual system perceives objects compares them. is invariant translation, rotation, scale change noise corruption. Moreover, allows accurate reconstruction boundary from feature vector used describe it. experimental results comparisons show good performance technique.