Documentation

OCR Support

Answer with photos of handwritten notes.

Installation

pip install learnlock[ocr]

Installs EasyOCR and Pillow.

Usage

During study, provide an image path instead of typing:

Challenge: How does backpropagation update weights?

> /home/user/notes/backprop.jpg

Extracting text from image...
Checking relevance...
Extracted: "Backpropagation uses chain rule..."

[Coach analyzes the extracted text]

Supported Formats

.png
.jpg / .jpeg
.webp
.bmp
.gif

OCR Engines

LearnLock tries in order:

EasyOCR - Better accuracy, works offline
Tesseract - Fallback (requires pytesseract)

Tips for Better OCR

Good lighting - avoid shadows
High contrast - dark ink on white paper
Clear handwriting - print style works best
Straight alignment - keep paper flat
Crop tightly - remove background

Troubleshooting

No OCR engine available

pip install learnlock[ocr]

Image content not related

The extracted text does not match the concept. Check the image or try clearer notes.

No text found

Try better lighting, higher resolution, or clearer handwriting.

Configuration