A statistical model for magnitudes and angles of wavelet frame coefficients and its application to texture retrieval
作者：
Highlights：
• We propose a new texture descriptor, based on wavelet frames.
• Magnitudes and angles of details are statistically modeled.
• A KLD-based distance gives relative importance to each model and resolution level.
• The distance is extended in a natural way to provide rotation invariance.
• Compared with other results, we obtain superior retrieval rates for a set of fewer parameters.
摘要
Highlights•We propose a new texture descriptor, based on wavelet frames.•Magnitudes and angles of details are statistically modeled.•A KLD-based distance gives relative importance to each model and resolution level.•The distance is extended in a natural way to provide rotation invariance.•Compared with other results, we obtain superior retrieval rates for a set of fewer parameters.
论文关键词：Image retrieval,Rotation invariant,Statistical models,Texture descriptor,Wavelet frames
论文评审过程：Received 17 June 2013, Revised 11 February 2014, Accepted 5 March 2014, Available online 15 March 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.03.004
