Abstract:
Modern information technology has evolved, and image processing's importance in
our lives is growing. There are many practical applications, such as digital
publishing, image databases, and imaging technology for medicine, a camera, a
visible wireless phone, etc. For better and finer images for users, visuals frequently
need to be zoomed in and out of or recreated in higher resolution from lower
resolution. An effective technique for enlarging an image is to interpolate the discrete
source image. Image interpolation techniques are directly applied in image zooming.
A zoom is simply a uniform scaling of the image. In the literature, numerous picture
interpolation techniques have been suggested.
In this thesis is introduced a novel zoom method based on sampling theory,
which is not the same as the techniques known by the names of ‘Sinc interpolation’,
‘zero-padding’ and ‘Fourier zoom’. Firstly, this zoom method is introduced in a
C/C++ program then in a Java program.
Many issues are raised by zooming, including the following: inescapable
effects of smoothing, restoration of high-frequency details without adding artifacts
and with the least amount of memory and processing time needed. Many efficient
zooming techniques are now in use. Additionally, new approaches were put out for
both efficient and effective algorithms.
