The super-resolution generative adversarial network (SRGAN) is a seminal work that is capable of generating realistic textures during single image super-resolution. However, the hallucinated details are often accompanied by unpleasant artifacts. To further enhance the visual quality, we propose a deep learning method for single image super-resolution (SR). Our method directly learns an end-to-end mapping between the low/high-resolution images. The method is based on depthwise separable convolution super-resolution generative adversarial network (DSCSRGAN). A new depthwise separable convolution dense block (DSC Dense Block) was designed for the generator network, which improved the ability to represent and extract image features, while greatly reducing the total amount of parameters. For the discriminator network, the batch normalization (BN) layer was discarded, and the problem of artifacts was reduced. A frequency energy similarity loss function was designed to constrain the generator network to generate better super-resolution images. Experiments on several different datasets showed that the peak signal-to-noise ratio (PSNR) was improved by more than 3 dB, structural similarity index (SSIM) was increased by 16%, and the total parameter was reduced to 42.8% compared with the original model. Combining various objective indicators and subjective visual evaluation, the algorithm was shown to generate richer image details, clearer texture, and lower complexity.

This paper presents an method of self-calibration of varying internal camera parameters that based on quasi-affine reconstruction. In a stratified approach to self-calibration, a projective reconstruction is obtained first and this is successively refined first to an affine and then to a Euclidean reconstruction. It has been observed that the difficult step is to obtain the affine reconstruction, or equivalently to locate the plane at infinity in the projective coordinate frame. So, a quasi-affine reconstruction is obtained first by image sequences, then we can obtain the infinite plane in the quasi-affine space, and equivalently to affine reconstruction. Then the infinite homography matrix can be calculated through the affine reconstruction, and then using the infinite homography matrix and constraints of the image of absolute conic to calculate the camera internal parameters matrix, and further to measure the metric rreconstruction. This method does not require a special scene constraints(such as prapllel, perpendicular) information, and also the camera movement informations(such as pure translation or orthogonal movement ), to achieve the goal of self-calibration. The theoretics analysis and experiments with real data demonstrate that this self-calibration method is available, stable and robust.