用于以增加的分辨率修复图像的系统和方法
By increasing the resolution and using super-resolution and thinning networks during the image restoration process, the problems of blurry and inconsistent restoration areas in existing technologies are solved, achieving high-resolution and refined image restoration effects.
CN116134477BActive Publication Date: 2026-07-17GOOGLE LLC
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GOOGLE LLC
- Filing Date
- 2021-10-14
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing image inpainting techniques, when using neural networks, suffer from problems such as blurred or faded restoration areas or inconsistencies with the original image. Furthermore, the restored images are not suitable for image processing operations such as edge detection or object detection.
Method used
After coarse restoration using a neural network, the image resolution is increased by using a super-resolution network as input, followed by thinning processing using a thinning network to achieve high-resolution image restoration.
Benefits of technology
It improves the resolution and detail clarity of the repaired area, making the repaired image compatible with surrounding features and suitable for image processing operations.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116134477B_ABST
Abstract
一种方法包括接收输入图像。该输入图像与待修复的一个或多个掩模区域相对应。该方法包括：向第一神经网络提供该输入图像。该第一神经网络输出第一分辨率的第一经修复图像，并且该一个或多个掩模区域在该第一经修复图像中被修复。该方法包括：通过将该第一经修复图像的分辨率从该第一分辨率增加到第二分辨率来创建第二经修复图像。该第二分辨率大于该第一分辨率，使得一个或多个经修复的掩模区域具有增加的分辨率。该方法包括：将该第二经修复图像提供给第二神经网络。该第二神经网络输出该第二分辨率的第一细化的经修复图像，并且该第一细化的经修复图像是该第二经修复图像的细化版本。
Need to check novelty before this filing date? Find Prior Art