Image Sharpening

Sharpness of a reproduced image is one of the important criteria for quality of reproduction. Due to influences from some external factors, images, which are transferred between devices and transformed between color spaces, are not only changed in their color and their gradation, but also blurred in their details. But the contours and edge information in image needs usually be extruded. Therefore, except for the color correction and gradation adjustment of image, the enhancement of image-sharpness, which is focused in this thesis, is a very important thing in image processing.In this paper, some conventional methods for image sharpening were compared each other. According to the principles for the image sharpening, a color and tone-level oriented sharpening method is researched. With this method, image details under different tone-levels and in different hue of colors can be sharpened with different strength.Programming on platform of Visual C++, following functions are realized-discrimination of tone ranges, hue-angle and hue-range; specialized setting of sharpening parameters for different target details. The target details can be enhanced specifically with USM (Unsharp Masking).