A novel super resolution approach for computed tomography images by inverse distance weighting method

In this study, a single image super-resolution approach, which is an integrated use of inverse distance weighting and histogram equalization methods, is proposed. It is aimed to reduce the detail loss which will be the result of increasing the dimensions of the images. In the proposed approach, while the edge information of the image is successfully preserved by the inverse distance weighting method, the brightness values of the pixels are approximated to the true image through general histogram equalization. The performance of the approach has been tested using a computed tomography database. The results obtained were compared in detail with various super-resolution methods available in the literature. When comparing the performance of the method, correlation coefficient, peak signal to noise ratio, structural similarity index and Pratt's figure of merit were used.