Super-resolution method using local binary pattern classification and linear mapping
AssigneeUniversity-Industry Cooperation Group of Kyung Hee University
InventorsSung-Ho Bae, Sungyoung Lee
Life of this patent
Estimating the term…
Abstract
A super-resolution method according to the inventive concept may include receiving an input image of low resolution, separating the input image into low resolution (LR) unit patches, classifying a texture type of each of pixels included in the input image using a local binary pattern for the LR unit patches, generating high resolution (HR) unit patches corresponding to each of the pixels based on a mapping kernel corresponding to the texture type, and combining the HR unit patches based on a predetermined setting to generate an output image of high resolution.
Claims15
A super-resolution method comprising:
- receiving an input image of low resolution;
- separating the input image into low resolution (LR) unit patches, each of which is set as a group of 3×3 pixels included in the input image;
- classifying a texture type of each of pixels included in the input image using a local binary pattern for the LR unit patches;
- arranging the LR unit patches in a specific order to generate a LR unit matrix having a size of 1×9:
- multiplying the LR unit matrix by a mapping kernel corresponding to the texture type, to calculate a high resolution (HR) unit matrix having a size of 4×1, wherein a size of the mapping kernel has a size of 9×4, which is determined depending on a size of the LR unit matrix and a size of the HR unit matrix;
- converting the HR unit matrix into HR unit patches, each of which is set as a group of 2×2 pixels;
- and
- combining the HR unit patches based on a predetermined setting to generate an output image of high resolution.
Dependencies are inferred from the claim text.