Image rain removal method and system, terminal and storage medium

Abstract

The invention provides an image rain removal method and system. The method comprises the following steps: constructing an image rain removal model based on a recursive residual cavity space pyramid pool network; constructing a training sample set; training the image rain removal model by adopting the training sample set to obtain a rain-free image generation model; and inputting the rain graphs of different sizes into the rain-free image generation model to obtain corresponding rain-removed images. Meanwhile, the invention provides a corresponding terminal and a storage medium. According to the invention, based on the recursive residual cavity space pyramid pool network, the problem of image restoration in rainy days is effectively solved; by extracting and fusing multi-scale information in a rain map, the quality of a recovered rain-free image is higher, and a long-short-term memory network module is introduced to enhance the dependency between stages; a mixed loss function is introduced, so that the details of the recovered image are finer, and the edge is clearer; the structural similarity and the peak signal-to-noise ratio of the rain-removed reconstructed image can be effectively improved, and a better effect is obtained subjectively.

Description

Technical field[0001]The present invention relates to the field of image processing and reconstruction techniques, and in particular, to an image retraction method, system, terminal, and storage medium based on a recursive residual hollow space pyramid network.Background technique[0002]With the development of science and technology, human society is entering the information society. The image has become an important source of human utilization information due to its large amount of information. In recent years, computer vision systems are widely used in various fields of society. In many algorithms of computer vision, such as image segmentation, target identification, behavior detection, etc., requires effective information in the image to be implemented. However, in the case where meteorological conditions are not good, if it rains, the image of the image has been collected by the outdoor imaging system, the image blur, color distortion, etc., directly affects the extraction of ima...

AI Technical Summary

Problems solved by technology

If this method is applied to image deraining technology, its target detection model pays more attention to the high-level semantic features and location information of the image, while image deraining pays more attention to the low-level features at the pixel level. It may b

Images