Image Reconstruction from Multiscale Singular Points Based on the Dual-Tree Complex Wavelet Transform

The representation of an image with several multiscale singular points has been the main concern in processing. Based on dual-tree complex wavelet transform (DT-CWT), a new reconstruction (IR) algorithm from is proposed. First, was transformed by DT-CWT, which provided multiresolution analysis. Then, accurate for IR were detected DT-CWT domain due to shift invariance and directional selectivity properties DT-CWT. Finally, images reconstructed phases magnitudes alternating orthogonal projections between CT-DWT space its affine space. Theoretical analysis experimental results show that proposed feasible, efficient, offers certain degree denoising. Furthermore, outperforms other classical algorithms terms performance metrics such as peak signal-to-noise ratio, mean squared error, structural similarity.