Adaptive Laplacian Filtering Interpolation Scaling Algorithm

In the field of video display, with the increases of resolution, the problems of transmission bandwidth, display adaptation and image display quality have gradually become prominent. To solve the above problems, we propose a new adaptive filtering interpolation scaling algorithm composed of an adaptive Laplacian filter and a bilinear interpolation scaling algorithm. The Laplacian filter alleviates the blurring effect of bilinear interpolation and enhances the image edges. Furthermore, an adaptive technology is used to enhance the effects of Laplacian filters. Bilinear interpolation was chosen for its simplicity of computation and good quality of scale. Experimental results show that compared with other methods, the average peak signal-to-noise ratio of the proposed is 0.64 dB higher than the bilinear interpolation scaling algorithm, and the calculation process is simpler.

I. Introduction

Digital image scaling refers to resizing a digital image based on the discrete pixels of the original image. It is essential to match the source of input video to the display resolution for image scaling technology. Image scaling is widely used in multimedia^{[1] - [2]}, consumer electronics^{[3] - [4]}, medical imaging^{[5] - [6]}, and other fields. For example, in multimedia video transmission, the scaled module is often required to alleviate the bandwidth limitation and ensure the quality of the scaled image. In consumer electronics, images should be enlarged or shrinked to fit different display devices. In medical images, images usually need to be magnified several times to observe subtle differences, which has higher requirements on the quality of the enlarged images. With the further development of display technology, high-resolution displays have gradually become a hot topic, but high-resolution displays are limited by data transmission bandwidth. Image scaling can effectively alleviate the problem of bandwidth walls. Therefore, image-scaling technology has become a research hotspot, and researchers have proposed many practical image-scaling algorithms.