Abstract:
Ultrasound imaging is a widely used in medical diagnosing imaging techniques.
Because of following: low cost, painless and considered harmless to the patients.
However, the main problem facing ultrasound images is speckle noise, which is
inherent multiplicative noise that reduces the overall quality of the images in term
of resolution and contrast.
Edge detection in medical image is an important task for object recognition of the
human organs, and it is an essential pre-processing step in medical image
segmentation and 3D reconstruction. Successful results of image analysis
extremely depend on edge detection. Up to now many edge detection methods
have been developed. But, they are sensitive to noise.
So, this thesis proposes, a new method of speckle noise reduction with ability to
enhancing image edges, i.e. a modified version of hybrid median filter, it works in
sub-windows and it is divided it vertical, horizontal and diagonal regions. Then it
computes the mean and the median value of diagonals element, and maximum of
vertical and horizontal elements in a moving window. After that the two values are
compared with the central pixel and the median value of the three values will
be the new pixel value. Ultimately six edge detection techniques were applied
using Proposed Method and hybrid median filter images.
The experimental result shows that by modifying the hybrid median filter, and
combining it with Roberts edge detection operator this gives better edge preserving
characteristics than hybrid median filter, edge detection helps in extracting useful
features for pattern recognition. and the proposed method is highly effective in
noise separation on compression other filters using image quality evaluation
metrics.
