Real-time color image filtering method based on nonlocal domain transformation

Abstract

The invention discloses a real-time color image filtering method based on nonlocal domain transformation. The method comprises the following steps: using a two-dimensional color image as a two-dimensional manifold on a five-dimensional vector space, and defining distance-weighted keeping transformation to obtain a multi-channel weighted isometric function; using recursive filtering to realize one-dimensional nonlocal domain transformation filtering of linear time; conducting separated one-dimensional nonlocal domain transformation filtering to the two-dimensional color image. The nonlocal domain transformation method can effectively involve nonlocal information into an image filtering algorithm. Under the premise that a small amount of computation is added, a high-quality filtering image is effectively obtained, and meanwhile the two aspects of requirements that the edge information in the image is not fuzzy are taken into account. A single color (or gray) image is adopted as an input source, the algorithm complexity is lower, the filtering effect is good, and a variety of computer vision application requirements can be met. The method is especially suitable for image detail enhancement, picture stylizing, image colorizing and pencil drawing conversion.

Description

technical field [0001] The invention belongs to the field of image information processing, in particular to a real-time color image filtering method based on non-local domain transformation. Background technique [0002] In the process of image generation and transmission, the image will inevitably be disturbed and affected by noise, which will reduce the quality of the image. How to filter the image smoothly and keep the edges and details of the image from being blurred has always been one of the hot spots in the field of digital image processing research. Image smoothing and filtering plays a very important role in image processing, and the filtered results have a direct impact on subsequent processing such as image edge detection, image segmentation, image enhancement, feature extraction, and image recognition. [0003] Image filtering has received extensive attention in the past two decades, and the most famous filtering methods are anisotropic diffusion and bilateral f...

AI Technical Summary

Problems solved by technology

However, anisotropic diffusion requires an iterative solution and bilateral filtering uses spatially varying weights, so it also has high computational complexity.

In recent years, scholars have also tried to improve anisotropic diffusion and bilateral filtering, for example: the weighted least squares framework proposed by Farbman et al., which needs to solve a sparse linear system; the boundary avoidance wavelet (EAW) proposed by Fattal et al. The calculation speed is fast, but there is a strict limit on the size of the smooth kernel

Due to the poor filtering effect or the high complexity of the algorithm, the above methods are rarely used in engineering

Moreover, these image filtering methods mainly focus on the processing of grayscale images. When processing color images, the three channels of R, G, and B are often processed separately, ignoring the integrity of color information, and the obtained results often have ringing. effect

Images