🔒
No file uploads
Your image, rendered PDF pages, and OCR text stay in browser memory.
PaddleOCR · Local processing
OCR & Editable PDF
Recognise, review, preview, and export entirely in your browser
✍️
Line-by-line editing
Use detected text boxes and confidence scores to review every OCR result.
📄
Searchable PDF
Rebuild visible text at its detected positions in a selectable, searchable PDF.
PP-OCRv5 · WebAssembly · ONNX Runtime
Browser OCR powered by PaddleOCR.js
This tool uses the official @paddleocr/paddleocr-js SDK to run the PP-OCRv5 text detection and recognition pipeline in your browser. It accepts an image or scanned PDF, finds text regions, recognizes each line, and lets you correct the result before creating a PDF.
How local OCR works
1
PP-OCRv5 pipeline
PaddleOCR.js combines text detection and text recognition. Each result includes the recognized text, polygon coordinates, and a confidence score.
2
WASM inference
The SDK manages OpenCV.js and ONNX Runtime. Inference runs through WebAssembly on your device without sending document pixels to an OCR server.
3
Model caching
The OCR model and WASM runtime are downloaded on first use and can be cached by the browser. Later runs are usually faster.
4
Private document workflow
Only public model and runtime resources are downloaded. Your selected file and corrected text are never uploaded by this tool.
Convert a scan into a searchable PDF
- 1
Choose one file
Open a JPG, PNG, WebP, BMP, or scanned PDF.
- 2
Automatic OCR
The file is rendered locally and recognition starts automatically.
- 3
Review the text
Edit any detected line in the page overlay or full-text view.
- 4
Export PDF
Download a clean PDF that reconstructs visible text using the detected layout.
OCR and searchable PDF FAQ
Does ShotEasy upload my document?
No. Document images, PDF pages, and recognized text remain in your browser. The browser only downloads the public OCR model, WASM runtime, and export font when needed.
Why is the first OCR run slower?
The first run initializes OpenCV.js and ONNX Runtime and downloads the PP-OCRv5 model. Browser caching can make later recognition faster.
Can I correct recognition mistakes?
Yes. Every detected text line is editable. You can edit it over the page or use the full-text panel before exporting.
What makes the exported PDF searchable?
ShotEasy writes corrected OCR lines as visible PDF text at their detected positions. The source image is not embedded, so compatible PDF viewers can search and select the reconstructed content.
Your privacy matters
Your photos are not uploaded. Editing happens in your browser.
Core ShotEasy Tools
Quick links to screenshot, image, video, and local browser tools.
EnglishPortuguês (Brasil)españolFrançaisDeutsch日本語English (India)हिन्दी한국어Bahasa IndonesiaTiếng ViệtУкраїнськаРусский简体中文
Copyright © 2025 ShotEasy