Light-weight image super-resolution reconstruction method and device based on frequency domain separation network

• Main Authors: SUN YISHI, GAO CHANGXIN, SANNON
• Format: Patent
• Language: Chinese; English
• Published: 10.03.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a light-weight image super-resolution reconstruction method and device based on a frequency domain separation network, and belongs to the technical field of image processing, and the method comprises the steps: S1, carrying out the shallow feature extraction of a to-be-processed low-resolution image through employing the convolution of a preset size, and obtaining a shallow feature; s2, performing deep feature extraction on the shallow features by using a Fourier residual module group based on a frequency domain separation network to obtain deep features corresponding to each module; s3, combining the deep features corresponding to each module, and performing feature fusion by using convolution to obtain fused features; and S4, combining the shallow layer information serving as residual information with the fusion features so as to generate a high-resolution image. According to the method, the Fourier residual module group corresponding to the frequency domain separation network is des