Image super-resolution reconstruction method of improved convolution network based on data enhancement

Abstract

The invention discloses an image super-resolution reconstruction method of an improved convolution network based on data enhancement. The method comprises steps of data enhancement and network structure improvement. On one hand, variety of samples are increased by adopting the mode of multi-angle rotation and rotation of a sample set, so features of multi-angle backgrounds can be acquired and rotation invariance of the features is achieved. Sufficient feature information is beneficial to improvement of reconstruction precision of an image. On the other hand, in the network model provided by the method, features are extracted by use of a deep convolution neural network, so multiple layers of convolution layers are beneficial to more advanced extraction and more complete features. De-convolution layers are used as reconstruction layers for carrying out feature mapping on output of the convolution layers so as to recover the image resolution, so the super-resolution image is obtained. The convolution layer is lack of the rotation invariance feature, so due to the variety of the samples in the provided method, an objective of increasing parameters is achieved, so the network is well fitted, improvement of the reconstruction precision is finally achieved and an effect of accelerating convergence speed of a network training is achieved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image super-resolution reconstruction method based on a data-enhanced improved convolutional network. Background technique [0002] In recent years, image super-resolution reconstruction technology has gradually matured and is widely used in the field of medical images, satellite images, and face recognition. The techniques can be divided into three categories: interpolation-based algorithms, reconstruction-based algorithms, and learning-based algorithms. Since learning-based algorithms are more effective, most scholars are exploring and researching on this basis. Currently, learning-based methods learn the mapping relationship between low-resolution image patches and high-resolution image patches. Although the sparse coding algorithm proposed by Yang et al. has some breakthroughs, this method mainly adjusts the process of dictionary learning and sparse regularizati...

AI Technical Summary

Problems solved by technology

However, the small data set of the network during the training process makes the convergence of the network training slow, and at the same time, the structure characteristics of the network layer cause the lack of useful information obtained, which ultimately leads to low image reconstruction accuracy.

[0003] CN106228512A is based on the learning rate adaptive convolutional neural network image super-resolution reconstruction method, which is directly applied to the convolutional neural network. The innovation is to add a BN layer after each convolutional layer, but only from the perspective of network convergence speed, and Does not combine image information with network structure

[0004] CN1062204449A A single image super-resolution reconstruction method based on a symmetrical deep network, by combining convolutional layers and deconvolutional layers, increasing the network depth at the same time, strengthening the reconstruction ability of image detail parts, and obtaining better image super-resolution High-rate reconstruction effect, but its reconstruction accuracy needs to be improved

Images