Edge pass filtering of the digitized image of G-banded human metaphasic chromosomes using Daubechies wavelets

• Main Authors: Rome, Michael P., Carbonell, Christopher Anthony G.
• Format: text
• Language: English
• Published: Animo Repository 1998
• Subjects: Human chromosomes; Image processing > Digital techniques; Electric filters (Wave-guide); Karyotypes; Mathematical analysis; Physics
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/4499
• Institution: De La Salle University

Two-dimensional wavelet transform and inverse wavelet transform methods are used to edge pass filter the digitized image of G-banded human metaphasic chromosomes. Several Daubechies' family of wavelets are applied in the study of improving the phase information of the digitized image of human chromosomes. Implementation of Wavelet Transform and Inverse Wavelet Transform is programmed in C++ programming language using the Pyramidal algorithm. Each digitized image of a G-banded human metaphasic chromosome is configured in a 256 grayscale pixel format and in a PCX graphics file format whose image size is delimited to 256 X 256 pixel size. Specimens imaged under a transmissive light microscope is captured by a CCD (Charged Coupled Device) camera and then digitized by a video frame grabber. The digitized image is stored and processed by a Pentium PC that runs the edge pass filtering program. Low phase band data of the transformed image is then set to zero to highlight the high phase band component. An Inverse Wavelet Transform is implemented to restore the high phase band information of the digitized G-banded human metaphasic chromosome. All Daubechies' family of wavelet filters can evidently increase the edge phase contrast. Higher Daubechies filters brighten up the background of the image while lower Daubechies filters darken the background of the image. There is no significant difference between images among different scales for the same wavelet filter.