Partial differential equation image amplification method based on angular point protection

Abstract

The invention discloses a partial differential equation image amplification method based on angular point protection. The partial differential equation image amplification method comprises the following steps: 1) inputting an image I, and calculating the definition R1 of the image I; 2) carrying out discretization on an image u; 3) calculating an angular point Cu of the image; 4) inputting the values of parameters k, [delta]t, [gamma] and a; 5) adaptively obtaining a fidelity term coefficient [lambda]; 6) for each iteration, obtaining the first-order partial derivatives ux and uy and the second-order partial derivatives uxx, uxy, uyy, u[eta][eta] and u[zeta][zeta] of each pixel of the image u; 7) on the basis of a discrete form of a directional partial differential equation model, obtaining a numerical solution ui,j<n+1> of each pixel of the image u; 8) calculating the definition Ru(n+1) of an evolution image u<n+1> obtained in the step 7); and 9) when a maximum iteration frequency judgment formula is met, taking the numerical solution ui,j<n+1> as an amplified image. The partial differential equation image amplification method can be widely applied to the field of digital image amplification.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to a partial differential equation image enlargement method based on corner point protection, which can be used in the image enlargement field. Background technique [0002] Image magnification is an image processing technology that obtains a high-resolution version of a low-resolution image through certain processing. Its core is how to supplement the gray value of the required pixels, that is, how to realize the supplement of image data. . Image magnification technology is widely used in the fields of public security, monitoring and remote sensing. [0003] At present, image interpolation enlargement method is an important direction of research and development in the field of image enlargement. It is a low physical resolution (Low Physical Resolution, referred to as LPR) fuzzy image interpolation reconstruction corresponding to high physical resolution (High Physical Reso...

AI Technical Summary

Problems solved by technology

Traditional partial differential equation image processing methods can achieve interpolation and magnification, but these methods do not impose specific constraints on the characteristics of the image, which will also lead to blurred edges and loss of details after magnification

Shao Wenze and others combined the directional diffusion PDE with the corner constraint algorithm (see literature 4), which suppressed the diffusion of the directional diffusion PDE in the corner area and the edge, thereby improving the visual effect of the image, but the internal diffusion of the barrier indicator function was completely prohibited. It will lead to excessive protection of the corners, making the connection between the corners and the edges unnatural, and affecting the image quality

[0005] In addition, the application of PDE to image processing involves a very important issue, that is, the selection of equation parameters

Images