Fractal Vision In Characterizing A Digitized Image

Characterization of a digitized image is often required in nondestructive evaluation. Apart from the existing image processing techniques, a recent method to analyze and quantify complex images has been developed based on the fractal theory. As reported, it is possible to get salient information from any digitized image by evaluating the fractal dimension, the defining characteristic of a fractal. In the present work fractal analysis is performed on digitized images obtained from of domains having inserts. The data, digitized by a digitizer board, is obtained from the image of the domain grabbed by a CCD camera. Analysis is performed to compute the fractal dimension of a digitized image, which represents the quality of the image in a quantitative sense. The results show that the fractal dimension is not only sensitive to the presence of a different region (insert) in an image, but is also sensitive to the relative position of the insert in the domain.