Image interpolation method and image interpolation apparatus

Abstract

The invention discloses an image interpolation method and an image interpolation apparatus. The image interpolation method comprises: edge point information of an image is obtained; according to the edge point information and peripheral reference points of a to-be-interpolated point, whether the to-be-interpolated point is located in an edge region is determined; if so, an angle class which the edge direction of the to-be-interpolated point belongs to is obtained, wherein edge directions of pixel points of the image are classified into m angle domains in advance and the m angle domains are divided into n angle classes in advance; on the basis of the angle class which the edge direction of the to-be-interpolated point belongs to, a corresponding one-dimensional horizontal interpolation kernel is used for carrying out horizontal interpolation on the to-be-interpolated point; and after horizontal interpolation completion, vertical interpolation is carried out on the to-be-interpolated point by using a preset one-dimensional vertical interpolation kernel. On the basis of the technical scheme, the image quality after interpolation processing can be guaranteed; the computation complexity during the interpolation process can be reduced; and software and hardware resources that may be occupied during the interpolation process can be reduced.

Description

technical field [0001] The invention relates to the field of display technology, in particular to an image interpolation method and an image interpolation device. Background technique [0002] In a video display system, due to the various resolutions of video sources, video input signals need to undergo image interpolation and scaling before being sent to a display device for display. Since the performance of image interpolation scaling directly determines the quality of video display, video display systems urgently need high-quality image interpolation scaling to improve users' visual experience. [0003] At present, some existing two-dimensional interpolation kernels such as Nearest interpolation kernel, Bilinear interpolation kernel, Bicubic interpolation kernel, Lanczos interpolation kernel (including: Lanczos1 interpolation kernel, Lanczos2 interpolation kernel, Lanczos4 interpolation kernel and Lanczos6 interpolation kernel) are often used in the prior art. Kernel to ...

AI Technical Summary

Problems solved by technology

[0004] However, when using these existing two-dimensional interpolation kernels for interpolation operations, when a smaller two-dimensional interpolation kernel is selected for operation, there is often obvious jaggedness or distortion in the final image; when a larger two-dimensional interpolation kernel is selected Although the image distortion can be avoided when the interpolation kernel performs calculations, the entire calculation process takes a long time and has higher requirements for hardware

Images