A Silk Pattern Image Restoration Method Based on Context Fusion
A repair method and context technology, applied in the field of image processing, can solve the problem of ignoring semantic features, and achieve the effect of overcoming the problem of color distortion and texture distortion.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment
[0027] figure 1 A flow chart of a context-integrated silk pattern image restoration method provided in this embodiment, specifically includes the following steps:
[0028] Step 1: Build a silk pattern image dataset. Collect 1000 intact silk pattern images to obtain a silk pattern image dataset. The spatial resolution of each silk pattern image is normalized to 224×224×3 by cropping and scaling (ratio of width and height). Randomly select 800 silk pattern images from the silk pattern image dataset to form the training set, and the remaining 200 silk pattern images form the test set. Using JPEG compression (quality factors set to 95, 90, 85, 80, 75, 70, 65, 60, 55, 50), rotation (rotation angles set to 90 degrees, 180 degrees, 270 degrees), and mirror flip (The flipping direction is set to horizontal and vertical respectively). The operation sequentially processes the silk pattern images in the training set, expands the number of training samples by 15 times, and obtains a tr...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


