Image editing method improves removing moiré patterns from screen photos
Dual-Scale Condition-Shift Editing for Robust Real-World Image Demoiréing
Computer Vision and Pattern Recognition
Summary
Photos taken of screens often have annoying wavy or striped patterns called moiré that make them look bad. The authors present a new way to clean up these photos by editing them in two steps: first fixing bigger color problems in a smaller version, then adding back fine details in the full-size image. They also created a new test set of tricky screen photos taken by phones to check how well their method works. Their approach outperforms previous ones by better removing these wavy patterns while keeping the original image content.
What this means in practice
- •For mobile app developers: Improve photo editing apps to better remove moiré patterns from screen capture photos automatically.$Commercial implications: Enables selling advanced camera apps or photo editors with superior screen photo cleanup features.
- •For display manufacturers: Analyze and reduce screen-camera interference effects by testing devices against challenging moiré patterns.
Authors
Shiyang Zhou, Haijin Zeng, Congpei Qiu, Zhengcen Li, Ping Chen, Xiangming Wang, Zheng Li, Yongyong Chen, Jingyong Su
Abstract
Robust demoiréing of screen-recaptured images remains challenging: moiré varies in frequency, color, and spatial support across display-camera pairs and capture conditions. Dedicated restoration networks trained on finite paired data struggle with this diversity, especially when low-frequency color banding and high-frequency stripes co-occur. We present DemoiréEdit, a content-preserving editing framework that adapts the natural-image prior of a pretrained flow-matching editor to demoiréing. First, an endpoint-supervised training objective supplements velocity matching with pixel, perceptual, and frequency-domain restoration losses. Second, a dual-scale condition-shift strategy performs global correction at low resolution, then refines the image at full resolution by shifting the denoising condition from the first-stage output to the original input for detail recovery. We also introduce DailyMoiré130, a benchmark of challenging moiré in everyday screen photography, comprising 130 smartphone captures across diverse displays with low-frequency color banding, fine high-frequency stripes, and mixed-frequency interference. On the challenging DailyMoiré130 benchmark, our method achieves the lowest residual moiré area and receives 71.3% overall-quality preference in a blind Best-of-7 study with 31 participants. We also evaluate on paired FHDMi/UHDM, reducing residual moiré confidence by 47.1%/9.3% and increasing MUSIQ by 4.3%/2.0% over the best baseline for each metric.