一种图片背景文字多语言自动替换方法
By constructing a foreground and background detection model based on convolutional neural networks and a text processing module, the problem of automatic replacement of background text in product images on e-commerce platforms was solved, enabling multilingual translation and lossless text removal, thus improving user experience and efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- FOCUS TECH
- Filing Date
- 2023-02-15
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies cannot effectively automate the replacement of background text in product images on e-commerce platforms with multiple languages, resulting in a poor user experience for users of less common languages, and the modification process is complex and costly.
We designed a foreground and background detection model based on convolutional neural networks, combined with text detection and recognition, language recognition and translation converter, to build a region text remover and achieve automatic multilingual replacement of background text.
It achieves automated multilingual replacement of background text, accurately identifies and translates background text, and removes text areas without loss, thereby improving user experience and reducing labor costs.
Smart Images
Figure CN116071757B_ABST