Screen Translator for Games

This tool processes on-screen text by accepting an image input (screenshot or frame) and converting visible text into a machine-readable form via OCR. It then translates the extracted content into a target language using neural machine translation, with special handling for gaming terminology to preserve intended meaning. The workflow avoids relying on UI data and focuses on the textual content found in menus, dialogues, item names, and tutorials. It outputs the translated text and an extraction confidence score to assist localization reviews or direct gameplay understanding.
Conceptually, OCR recognizes characters, segments blocks of text, normalizes fonts, and applies layout-aware alignment to improve translation accuracy. The translation module uses a glossary-backed model, aligns phrases, and selects gaming-appropriate equivalents. Post-processing preserves line breaks, UI rhythm, and screen-space constraints so results remain legible. Optional features include batch processing, user-provided glossaries, and language-pair customization to handle regional variants common in gaming communities.
Audience and differentiators: gamers who play multilingual releases, testers validating localization quality, speedrunners evaluating understandability, and content creators who subtitle streams. The tool’s differentiators are domain-aware glossaries for popular game genres, font-robust OCR with fallback for decorative text, and a focus on preserving UI structure in translated output. Typical use cases include translating JRPG menus, ARPG item descriptions, and in-game tutorials during live play or quick play sessions.
Core features include image input, OCR extraction, language pairing, glossary integration, and translation output, with optional report generation and export options. The tool is designed to integrate with localization pipelines or assist solo players. It does not render overlays on the live game; instead, it provides ready-to-use text suitable for translation memory or in-house localization workflows.