Image OCR

Extract text from images using Tesseract OCR.

Prerequisites

Tesseract OCR must be installed:
```
# Windows (via chocolatey)
choco install tesseract

# Or download from: https://github.com/UB-Mannheim/tesseract/wiki

# Verify
tesseract --version
```
Python packages:
```
pip install pytesseract Pillow
```

Quick Start

# Extract text from single image
python scripts/image_ocr.py image.png

# Output to file
python scripts/image_ocr.py image.png -o output.txt

# JSON output with confidence
python scripts/image_ocr.py image.png --json

# Batch process directory
python scripts/image_ocr.py --dir /path/to/images -o ./extracted

# Specify language
python scripts/image_ocr.py image.png --lang eng+fra

Output

Human-readable:

File: screenshot.png
Characters extracted: 1523
Confidence: 87.5%

[Extracted text appears here...]

JSON:

{
  "source": "screenshot.png",
  "text": "Extracted text content...",
  "char_count": 1523,
  "confidence": 87.5,
  "language": "eng"
}

Supported Formats

Format	Extensions
PNG	.png
JPEG	.jpg, .jpeg
GIF	.gif
BMP	.bmp
TIFF	.tiff, .tif

CLI Reference

python scripts/image_ocr.py IMAGE [OPTIONS]

Arguments:
  IMAGE                 Image file to process

Options:
  -o, --output PATH     Output file (default: stdout)
  -d, --dir PATH        Process directory of images
  -j, --json            Output as JSON
  -l, --lang LANG       Tesseract language code (default: eng)
  --dpi DPI             Image DPI for processing (default: 300)
  --psm MODE            Page segmentation mode (default: 3)
  -q, --quiet           Suppress progress output

Integration Notes

Called by file-router when image files are detected. Output text can be passed to quiz-generator or stored in __cc_validated_files/.

image-ocr

Image OCR

Prerequisites

Quick Start

Output

Supported Formats

CLI Reference

Integration Notes

Is this your skill?