Rain and snow removing method for image based on multiple guided filtering

Abstract

The invention discloses a rain and snow removing method for an image based on multiple guided filtering, relates to image processing, and provides a rain and snow removing method for the image based on the multiple guided filtering, which can improve the visual effect of the image captured under the rainy and snowy weather conditions, can improve the accuracy of a visual algorithm. The method comprises the following steps: 1, the image itself serves as the guided image, and the guided filter is performed to the image with rain or snow so as to obtain the high frequency part and the low frequency part; 2, the low frequency part with the enhanced edge serves as the guided image, and after the guided filter is performed to the high frequency part, a recovery image is obtained by adding the low frequency part; 3, the least value of the recovery image and an input image is taken so as to obtain a clear recovery image; 4, the balance in the clear recovery image and the recovery is taken so as to obtain the final recovery image with better definition and removal effect. The rain and snow removal of the single image is realized based on the difference of rain lines or snow flowers and background edges.

Description

technical field [0001] The invention relates to image processing, in particular to an image rain and snow removal method based on multiple guided filtering. Background technique [0002] For photos taken on rainy or snowy days, some bright rain lines or snowflakes are usually superimposed on the background image. These rain lines and snowflakes not only affect our visual effects, but also reduce the performance of some computer vision algorithms, such as objects Recognition, Tracking, Image Retrieval and more. Removing rain and snow has very good applications in video surveillance, urban traffic, and image retrieval. [0003] In the field of computer vision, the removal of rain and snow in videos has achieved good results, but with moving objects in the background, some algorithms are less effective. There are few ways to remove rain and snow from a single image. Yu-Hsiang Fu et al. + ",in Proc.IEEE Int.Conf.Acoustics,Speech&Signal Processing,May2011,Prague,Czech Republic...

AI Technical Summary

Problems solved by technology

[0003] In the field of computer vision, the removal of rain and snow in videos has achieved good results, but with moving objects in the background, some algorithms are less effective

There are few ways to remove rain and snow from a single image. Yu-Hsiang Fu et al. + ",in Proc.IEEE Int.Conf.Acoustics,Speech&Signal Processing,May2011,Prague,Czech Republic,pp.1453-1456) proposed a method for removing rain lines based on sparse coding and dictionary learning, although this method can remove rain well , but cannot remove the snow, and the algorithm is relatively complicated; Jing Xu et al. (Jing Xu, Wei Zhao, Peng Liu&Xianglong Tang, "An Improved Guidance Image Based Method to Remove Rain and Snow in a Single Image", CCSE Computer and Information Science Journal, 2012, vol.5, no.3, pp.49-55) proposed a rain and snow removal method based on the rain and snow imaging model that also uses guided filtering. Although the algorithm is simple, the removal effect is poor and blurred

Images