Wavelet processing of topographic high dynamic range images

For a 6H-SiC single crystal, experimental results of the digital processing of topographic images with various dynamic ranges are given. In this case, wavelet analysis was used. The images were obtained by x-ray topography based on the Borrmann effect and were presented in 8-, 16-, and 32-bit formats. It is shown that, on eliminating the experimental contrast granularity, 32-bit images turn out to be more convenient for decoding and contain more useful information about structure defects of a single crystal under study as compared to the 8- and 16-bit images. The wavelet analysis efficiency increases considerably for the digital processing of separate topogram fragments and depends on the choice of the region and the size of the reference image.