Improvement of image sharpness by histogram equalization through two-dimensional wavelet transform

Abstract

This thesis emphasizes the improvement approaches of quality on digital data to be more sharpness by two- dimensional technique. Today, the availability from the improvement of Digital Image Processing is used widely in various occupations such as medical field with reading the results from CT scanning, military field with the improvement of aerial map for country prevention, agricultural field with leaf analysis of its disease, and the educational field with the image processing and robot vision to increase its effective learning and remembering development. It can be seen that the image development through its sharpness is really crucial, therefore this thesis used the Gray Scale for transforming by two-dimensional wavelet and LL Band got from this wavelet transform to be improved by histogram technique. The histogram was divided into 4 parts based on its cumulative density at 0.25, 0.5, 0.75, and 1, then smoothed the data by separation, after that brought the results to invert wavelet which gives the new image through improved LL band. The research result revealed that the mean of the Absolute Mean Brightness Error (AMBE) was decreased while enable the image sharpness to suitably increase 80%.