A face image deblurring method based on adaptive mesh deformation

Abstract

The invention provides a face image deblurring method based on adaptive mesh deformation to improve the effect of face image deblurring. The face image deblurring method comprises the following steps:A, pre-processing the original face image, obtaining the blur kernel value of the original face image, and estimating the blur degree of the edge of the original face image; B, generating an adaptivetriangular mesh by using the preprocessed face image; C, construct a triangular mesh based on a radial basis function by using an adaptive triangular mesh; D, design a displacement function, and moving pixel near that edge of the face image to the edge according to the displacement function; E: carry out edge enhancement on that image obtained in the step D by use the pixel mesh conversion, and inserting the image after edge enhancement from the deformed mesh into the target uniform mesh to obtain the deblur face image. The above method keeps the texture of the image and makes the edge clearer, which has a good effect on the basic deblurring of the face image.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a face image deblurring method based on adaptive grid deformation. Background technique [0002] As an important means for people to obtain information, images play a very important role in both the scientific field and real life. In recent years, with the rapid development and progress of science and technology, many kinds of portable camera equipment have been popularized. Whether it is a hand-held digital camera specially used for taking pictures, or a smart phone that almost everyone owns in life, it makes it easier for people to record moments in life. . However, in the actual shooting and dissemination of images, due to the influence of the resolution of the machine, the lighting conditions of the imaging and the relative movement of the camera, etc., the information of the actually obtained face images is usually lost and the quality is degraded, whic...

AI Technical Summary

Problems solved by technology

The traditional method of non-blind optical blur removal for face images is to reversely obtain a clear image on the premise that the blur kernel is known, but in most practical applications, the optical blur kernel of the image cannot be known in advance, and noise and other factors The interference of the face image makes the optical blurring problem more difficult, which is also the challenge of the non-blind optical blurring of the face image.

Images