Screen shot image moire removing method based on convolutional neural network AMNet

Abstract

The invention relates to the field of computer vision, and aims to remove moire and output a corresponding clean picture. The invention discloses a screen shot image moire removing method based on a convolutional neural network AMNet, and the method comprises the following steps: analyzing moire features, and building a moire model in a screen shot image; building a data set; designing a network framework; designing a network structure; setting the learning rate of the network and the weight of the loss function of each part, training the convolutional neural network by using a deep learning framework Tensorflow until the loss converges, and generating a training model; and inputting the test picture with the moire pattern into a network to obtain a corresponding clean picture without themoire pattern. The method is mainly applied to occasions for removing the screen moire patterns.

Description

technical field [0001] The invention belongs to the field of computer vision, and relates to a method for removing moiré in screen shot images. Specifically, through a learning-based approach, pairs of moiré-patterned images and clean images directly downloaded or screenshots are used as input. Through training, the network learns the mapping relationship from moiré images to clean images. Finally, the process of inputting a picture with moiré pattern, removing the moiré pattern by the network, and outputting a corresponding clean picture is realized. Background technique [0002] Moiré refers to the irregular fringes produced when two arrays of different frequencies are aliased together. When a digital camera is used to shoot a digital display device, aliasing occurs between the photosensitive element of the camera and the display device, and it is easy to appear moiré, which seriously affects the quality of the captured image. Due to its irregular shape, various colors,...

AI Technical Summary

Problems solved by technology

However, in the existing research, there are few data sets dedicated to the moiré removal task, and in these data sets, there is no brightness difference between the moiré image and the corresponding clean image, and it is difficult for the network to pass the training of these data sets. Brightness information is learned, and real-world screen shots are often darker than clean images seen by the human eye

Therefore, the network trained with this data set cannot restore image brightness while removing moiré

Images