An image-guided filtering method and computing device

Abstract

The invention discloses a method for generating a guidance map in guidance filtering and a method for performing guidance filtering processing on an image to be processed based on the guidance map. The image guidance filtering method includes the steps of: executing a method for generating a guidance map on the image to be processed to generate gradient enhancement. image and the second guide map; calculate the maximum and minimum values ​​of the pixel values ​​in the second guide map; perform block processing on the second guide map, and calculate the standard deviation value of the pixels in each block; Perform offset processing to obtain a second guide map after offset processing; generate an adjustment coefficient of the guide filter based on the gradient enhancement image; and generate a certain channel in the image to be processed based on the second guide map and adjustment coefficient after offset processing. The image is subjected to guided filtering to produce a noise-removed image. The invention also discloses a computing device for executing the above method.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image-guided filtering method and computing equipment. Background technique [0002] In the field of image processing technology, due to the influence of shooting environment, camera, etc., the captured image often contains noise, which affects the image quality. Therefore, it is often necessary to filter the image to remove the noise in the image. However, in the process of processing, filtering often leads to excessive blurring, which causes the loss of image edge details. For example, if there are scattered hairs in the skin of a portrait, if the filtering method is not selected properly, the effect of skin and hair adhesion will be caused, and the original image will be lost. detail features. [0003] Commonly used filtering methods such as Wiener filtering, Gaussian filtering, etc., will cause excessive blurring in image noise reduction processing, especially th...

AI Technical Summary

Problems solved by technology

However, in the process of processing, filtering often leads to excessive blurring, which causes the loss of image edge details. For example, if there are scattered hairs in the skin of a portrait, if the filtering method is not selected properly, the effect of skin and hair adhesion will be caused, and the original image will be lost. the details of the features

[0003] Commonly used filtering methods such as Wiener filtering, Gaussian filtering, etc., in image noise reduction processing, will lead to excessive blurring, especially the loss of image edge details

The guided filtering method has been widely used in image filtering processing because of its edge preservation and high calculation efficiency, but the guided filtering still has the following problems in the filtering process: (1) using fixed filter coefficients, for high-frequency images in color images Noise, guided filtering must be filtered cleanly, and reasonable filter coefficients need to be artificially selected to filter out high-frequency noise, but it will cause excessive blurring of the edge area of ​​​​the image, and the effect of edge preservation cannot be achieved; (2) guided filtering depends on the guided image Reasonable construction, if the guide map cannot maintain edge consistency with the original image, it will also cause abrupt or overly blurred areas in the filtered result map, and lose the edge preservation effect; (3) Noise and edges belong to areas with large gradients , if the noise and the edge area are treated equally, the edge-preserving effect will also be lost. When the guided filter is applied, it needs to use the differential information to construct a guided map for edge-preserving filtering, and the differential information is easily affected by noise points, which needs to be effectively processed.

Images