打开 PDFnoted 网站（pdfnoted.com），点击「上传文件」按钮，选择你的扫描件PDF（如合同、发票、旧书页等）。PDFnoted 支持 JPG/PNG 内嵌的PDF及纯图像PDF，自动检测页面类型。
上传后，PDFnoted 自动进入OCR处理界面。点击「启用OCR」按钮，在弹出菜单中选择文档语言（如中文简体、英文、日文等）。多语言混合文档建议选‘自动检测’，PDFnoted 会逐页智能识别。
OCR完成后，PDFnoted 会生成带隐藏文字层的新PDF：原文档外观不变，但全文可选中、复制、Ctrl+F搜索。点击「下载PDF」获取可搜索版本；或点击「导出文本」一键保存为TXT，保留段落结构与换行。
Go to pdfnoted.com, click ‘Upload File’, and select your scanned PDF (e.g., contracts, invoices, book scans). PDFnoted supports image-heavy PDFs and embedded JPG/PNG pages — it auto-detects page type before OCR.
After upload, PDFnoted opens the OCR interface. Click ‘Enable OCR’, then choose document language (e.g., Simplified Chinese, English, Japanese). For mixed-language docs, select ‘Auto-Detect’ — PDFnoted performs per-page intelligent recognition.
Once OCR finishes, PDFnoted generates a new PDF with invisible text layer — visual fidelity preserved, but now fully searchable, copyable, and Ctrl+F friendly. Click ‘Download PDF’ for the searchable version, or ‘Export Text’ to save clean, line-break-aware TXT.
支持简体/繁体中文，基于深度学习模型优化。在清晰扫描件（分辨率≥200dpi）上，中文识别准确率达97%+；模糊或倾斜文档建议先用PDFnoted「增强图像」预处理。
可以。PDFnoted 在OCR完成后提供「优化PDF」选项，自动移除冗余图像元数据、压缩背景图层，同时保留文字层完整性和搜索功能，通常减小20–40%体积。
PDFnoted 免费版支持单次最多20页PDF的OCR识别，无水印、不限次数；高级版解锁批量处理、多语言并发识别及API接入，适合高频办公用户。
Yes — Simplified & Traditional Chinese are fully supported, powered by fine-tuned deep learning models. Accuracy exceeds 97% on clear scans (≥200 DPI); for blurry or skewed pages, use PDFnoted’s ‘Enhance Image’ tool first.
Yes. After OCR, click ‘Optimize PDF’ in PDFnoted — it removes redundant image metadata, compresses background layers, and preserves full text layer integrity and searchability, typically reducing file size by 20–40%.
Yes — the free plan allows up to 20 pages per OCR job, no watermark, unlimited jobs. Pro users get batch processing, concurrent multi-language OCR, and API access — ideal for teams handling dozens of scanned docs weekly.