图像修复方法、装置、电子设备和存储介质
By downsampling and instigating high-resolution images and generating low-resolution images for restoration, the problem of high computational resource and time consumption in existing technologies is solved, and efficient image restoration results are achieved.
CN116957985BActive Publication Date: 2026-07-17BEIJING BAIDU NETCOM SCI & TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING BAIDU NETCOM SCI & TECH CO LTD
- Filing Date
- 2023-08-10
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing high-resolution image restoration methods require significant computational resources and time, limiting their efficiency and real-time performance in practical applications.
Method used
A low-resolution image is generated by downsampling the image to be repaired, and then the image is repaired using a repair model to generate an intermediate image containing the area to be repaired. Finally, the repaired target image is generated based on the intermediate image and the original image.
Benefits of technology
It reduces the amount of computation, speeds up the repair process, saves computing resources, improves the repair effect, and avoids the loss of details.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116957985B_ABST
Abstract
本申请公开了图像修复方法、装置、电子设备和存储介质，涉及图像处理技术领域，尤其涉及算机视觉、深度学习等人工智能领域。方案为：获取待修复的第一图像；对第一图像进行下采样，以获取第二图像；对第二图像进行修复，以获取第三图像；根据第二图像与第三图像，生成第四图像，其中，第四图像中包括第二图像的待修复区域；根据第一图像和第四图像，生成已修复的目标图像。由此，通过对待修复的原图像进行下采样，得到待修复的低分辨率图像，对低分辨率图像进行修复，可以减少计算量，加速修复过程，节省计算资源，并且利用第四图像对第一图像进行修复，得到第一图像的修复图像，在对原图像完成修复的同时，可以进一步减少计算量，加速修复过程。
Need to check novelty before this filing date? Find Prior Art