Image → Text OCR
Client-side • No server • Light Saffron Theme
Tip: clearer image = better OCR

1) Add Image Idle

Upload an image or use your camera. Then optionally preprocess and run OCR.
Drag & drop an image here, or use the buttons. Recommended: well-lit, straight, sharp text.
Preview (Original)
Preview will appear here
Processed (Used for OCR)
If OCR is messy: try Grayscale + adjust Threshold, or increase Scale.
Language
Hindi OCR needs extra data download (first time). Keep internet on once.
Page Segmentation Mode (PSM)
Preprocess Mode
Threshold (only for B/W)
Current: 160
Scale (upscale before OCR)
Sharpen
OCR Progress
Waiting…
Help & Tips
How to use:
  1. Click Pick Image or Camera Capture (or drag-drop).
  2. Choose Language and optional Preprocess.
  3. Click Run OCR → text appears on the right.
  4. Use Copy / Download for the extracted text.
Best results:
  • Bright, sharp image; avoid shadows and glare.
  • Keep text straight (use Rotate).
  • Try B/W Threshold for printed text; try Grayscale for mixed backgrounds.
  • Increase Scale to 2× if text is small (slower).
  • For Hindi: choose hin or eng+hin. First run may download language data.
Privacy: Everything runs in your browser (no server upload). Language data loads from CDN.

2) Result

Extracted text appears below. You can edit it before copying/downloading.
Pro tip: If you want just numbers, or just one line, change PSM (e.g., single line = PSM 7).
Advanced
Optional: Replace common OCR mistakes
Example: O↔0, I↔1, extra spaces, weird quotes
Find & Replace (quick)
Works on the output box only