Image local structure self-adaption recovery method based on non-continuity indicator

Abstract

The invention provides an image local structure self-adaption recovery method based on a non-continuity indicator. The image local structure self-adaption recovery method based on the non-continuity indicator comprises the first step that a read image is initialized and an input RGB image is converted to a YCbCr color space; the second step that a trilateral scatter matrix is built and robustness to noise is improved; the third step that the non-continuity indicator is built to represent image local structure characteristics dynamically; the fourth step that an image degradation model is built; the fifth step that an image recovery optimization model is built according to the built non-continuity indicator, and the built model depends on the image local structure characteristics continuously; the sixth step that the recovery optimization model is resolved through a variational method, a gradient descent flow corresponding to the optimization model is obtained and is discretized through the half-point format, and an optimal recovery image is obtained. According to the image local structure self-adaption recovery method based on the non-continuity indicator, the recovery process can be controlled in a self-adaption mode according to the image local structure characteristics, more detailed structures of the image can be recovered, and the quality of the image is improved remarkably.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to a self-adaptive restoration method of image local structure based on a discontinuity indicator. Background technique [0002] Image restoration technology is a key issue in image processing and low-level vision. It is the basis for subsequent pattern recognition and high-level understanding. It has a wide range of application requirements. This technology can be applied to many fields such as traffic monitoring, military, and medicine. In terms of monitoring, due to the low resolution of the camera and the poor shooting environment, the quality of the captured image is degraded, and it is difficult to obtain the required detailed features from the image, such as the license plate information of the vehicle, which is not conducive to machine recognition or human identification. Therefore, improving image quality through image restoration technology has important theoret...

AI Technical Summary

Problems solved by technology

Through the research on the existing restoration technology, it is found that the existing methods are all controlled by the gradient, which cannot effectively distinguish the edge points from the noise points, so that they cannot remove the noise while protecting the edge structure of the image.

Images