Admissible Diffusion Wavelets and Their Applications in Space-Frequency Processing

摘要: As signal processing tools, diffusion wavelets and biorthogonal have been propelled by recent research in mathematics. They employ as a smoothing scaling process to empower multiscale analysis. However, their applications graphics visualization are overshadowed nonadmissible expensive computation. In this paper, our motivation is broaden the application scope space-frequency of shape geometry scalar fields. We propose admissible (ADW) on meshed surfaces point clouds. The ADW constructed bottom-up manner that starts from local operator high frequency, dilates its dyadic powers low frequencies. By relieving orthogonality enforcing normalization, locally supported admissible, hence facilitating data analysis processing. define novel rapid reconstruction, which recovers multiple bands frequencies low-frequency base full resolution. It enables operations localized both space frequency manipulating wavelet coefficients through filters. This paper aims build common theoretic foundation for host applications, including saliency visualization, feature extraction, spectral processing, etc.