Image enhancement method based on deep learning

Abstract

The invention discloses an image enhancement method based on deep learning. The method comprises the following steps: collection and processing of an original image, augmentation processing of the original image, sharpening processing of a target image, construction of a training set, construction of a data set and removal of class imbalance in the data set, removal of redundancy in the training set by wavelet transform, and training of the data set and the training set by a convolutional neural network. According to the invention, redundant information in the original image can be effectivelyremoved through gray processing of the original image; the redundancy is removed through wavelet transform, so that the training precision and effectiveness of the original image serving as a training set in the later stage can be comprehensively and effectively improved; in addition, in the process, the image can be denoised in real time, the image with a good visual effect can be obtained, theclass imbalance phenomenon is removed by adopting an interpolation-based SMOTE method, the problem of sample class imbalance can be effectively solved, and the enhancement effect of the image is ensured to be stable.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an image enhancement method based on deep learning. Background technique [0002] As an effective information carrier, image is the main source of information acquisition and exchange for human beings. Accordingly, the application field of image processing must involve all aspects of human life and work. Image enhancement is the low-level processing of images. preprocessing stage. But it is an important part of image processing, which plays an important role in linking the past and the future in the whole image processing process, and is crucial to the success or failure of high-level image processing. Its purpose is to improve the quality and visual effect of the image, or convert the image Into a form that is more suitable for human eye observation or machine analysis and recognition, so as to obtain more useful information from the image. [0003] The transmission of images i...

AI Technical Summary

Problems solved by technology

[0003] The transmission of images is inseparable from the transmission of the network. Due to the limitation of bandwidth, images must be compressed during transmission. The essence of compression is to weaken the high-frequency components of images, which will inevitably lead to blurred images. , there will be some weak textures near the edge of the image, and the overall feeling of the image is dirty. At the same time, in natural scenes, due to the limitation of the dynamic range of the image itself, lighting conditions, image capture equipment or the photographer's own technical level, and the image itself If the characteristics are modified to a certain extent, it will bring difficulties to the later image foreground and background segmentation, target recognition, target tracking and final image understanding and predictive analysis. Therefore, the present invention proposes an image enhancement method based on deep learning. To solve the deficiencies in the existing technology

Images