A method and device for removing rain from a single image

Abstract

The present invention provides a method and device for removing rain from a single image. Firstly, the pure rain region in the image to be processed is identified, and the pure rain region is further used as an input for dictionary learning and sparse representation to obtain a rainy dictionary. Furthermore, the image to be processed is sparsely reconstructed based on the rainy dictionary to obtain a rain mark mask. Finally, the rain removal process is performed on the image to be processed through the rain mark mask to obtain the target image, that is, an image without rain marks. Since this method only processes pixels with rain, the part of pixels without rain is kept. Therefore, the method in this paper has fewer rain marks left in the image after deraining, and maintains richer edges and textures in the target image, greatly improving the accuracy of deraining.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method and device for removing rain from a single image. Background technique [0002] Affected by rainy weather conditions, part of the texture and detail information of the image captured by the outdoor lens is easily blocked by the rain line, resulting in problems such as excessively bright local areas and blurred background images. The degradation of image quality in rainy days greatly restricts the functions of outdoor intelligent vision systems such as visual surveillance, visual navigation and target tracking. Moreover, the state of raindrop particles is changeable, and the direction and thickness of rainlines are different in different situations. Therefore, it is of high research and application value to study how to recover high-quality images from various rainy degraded images. [0003] Image deraining includes video image deraining and single image deraining. Diffe...

AI Technical Summary

Problems solved by technology

[0008] The existing image rain removal algorithm has achieved certain results, but there are still the following problems: (1) There is overlap between the rain line and the background image texture in the rainy image, and it is easy to mistake the texture similar to the rain line structure in the background image; If it is judged as rain lines, the image after rain removal will be over-smoothed; (2) if the rain lines are judged as background images, the rain removal will not be complete; the image without rain and the image with rain layer use the same dictionary for sparse representation, and there is no guarantee Complete separation of rain atoms and no rain atoms, easy to produce handmade edges after sparse reconstruction

[0009] In summary, the existing image deraining methods cannot effectively distinguish between rainy pixels and non-rainy pixels, and cannot take into account the advantages of fast image restoration and high quality restored images, which have certain limitations

Images