Image Scaling Algorithm Based on Elliptic Curve

Kai Li,Ke Ma,Shexiang Ma
DOI: https://doi.org/10.1109/icicas48597.2019.00154
2019-01-01
Abstract:When Digital Video image transformation from general screen (i.e., 4:3, etc.) to widescreen (i.e., 16:9, etc.), the length-width ratio of source image will change. So, it is difficult to adjust image resolution and convert the format. In this paper, a method of nonlinear scaling is proposed, aiming at overcoming the image distortion problems caused by conventional transform methods. The proposed method is that the vertical direction is transformed by linear interpolation, and nonlinear scaling transformation is conducted in the horizontal direction. The elliptic curve scaling function is used on both sides of the image and appropriate linear scaling factor is added in the central area, achieving the overall image of the nonlinear scaling factor change continuously. It can greatly reduce the image distortion in human vision. The simulation result shows that the distortion of the scaled image is acceptable, the proposed method is more efficient than conventional methods.
What problem does this paper attempt to address?