Image deraining method, device, readable storage medium and terminal equipment

Abstract

The present invention relates to the technical field of image processing, in particular to an image rain removal method, device, readable storage medium and terminal equipment. The method provided by the present invention includes: acquiring a rainy image; inputting the rainy image into the trained confrontational network model to obtain a clear image of the rainy image output by the generated confrontational network model after removing the rain; wherein, the generated confrontational network model includes The generative model and the discriminative model are set as adversarial training. The generative model is a fully convolutional network model obtained by using the features extracted from the second image to perform feature supervision training on the first image. The first image is the first training belt rain image, the second image is the first training clear image corresponding to the first training rain image, so as to perform the training of the generation confrontation network model through the feature supervision of the clear image, thereby optimizing the feature extraction function of the generation confrontation network model and improving the ability of the generation confrontation network model. The extraction accuracy of the original texture features of the image improves the rain removal effect of the image.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image rain removal method, device, computer-readable storage medium and terminal equipment. Background technique [0002] With the rapid development of image processing technology, target detection and recognition based on image analysis is widely used in various fields, such as automatic driving, target retrieval and other fields. Imaging blur and information coverage affect target detection and recognition, so it is necessary to remove rain from images with rain. [0003] At present, methods such as dictionary learning, non-local mean filtering, low-rank matrix decomposition, and Gaussian mixture model are usually used to remove rain. Remove the original texture information of the image, resulting in an over-smooth background, resulting in poor rain removal effect. Contents of the invention [0004] Embodiments of the present invention provide an image d...

AI Technical Summary

Problems solved by technology

[0002] With the rapid development of image processing technology, target detection and recognition based on image analysis is widely used in various fields, such as automatic driving, target retrieval and other fields. Imaging blur and information coverage affect target detection and recognition, so it is necessary to remove rain from images with rain

[0003] At present, methods such as dictionary learning, non-local mean filtering, low-rank matrix decomposition, and Gaussian mixture model are usually used to remove rain. Remove the original texture information of the image, resulting in an over-smooth background, resulting in poor rain removal effect

Images