A novel image sharpening technique based on 2D-DWT and image fusion

This paper presents a new image sharpening technique that can be applied to both gray-scale and color images. The proposed approach uses a multi-scale scheme and a wavelet based fusion algorithm. Specifically, the input image is initially processed by a cluster of un-sharp filters with different variance. The final image is then obtained with the aid of wavelet fusion. The application of the unsharp filters with different size Gaussian filters emphasizes important information in different frequency bands. It is shown that the proposed technique can be used as a preprocessing stage to general image fusion approaches. The quality of the resulting images is evaluated using three different sharpening indices. The experimental investigation demonstrates the effectiveness of the proposed technique.