Super-resolution of ship infrared-polarized image based on GAN and DWT
Aiming at the problem of low resolution and unclear details of infrared polarization imaging of sea ships,a method combining wavelet transform and generative adversarial network is proposed to improve image resolution.Firstly,the pure convolutional neural network model(ConvNeXt)is used to improve the super-resolution network(SRGAN),and the original low-resolution ship infrared polarimetric image is denoised by using non-local mean.Then,the low-resolution image is initially super-resolved with the improved SRGAN,and the detail information of the initial super-resolved image is extracted using a two-dimensional discrete wavelet transform.Finally,the detail information is fused with the original low-resolution ship infrared polarization image through the inverse wavelet trans-form.Compared with the traditional super-resolution method,the peak signal-to-noise ratio and structural similarity of the super-resolution image obtained by the proposed method are significantly improved.In this paper,the infrared po-larization image super-resolution and detail information fusion isachieved at the same time and the obtained super-res-olution image not only retains the infrared polarization information of the original image,but also fuses the high-resolu-tion detail information.
万方数据
维普
