Business Card Scanner

Script: scan_cards.py

Description

This script splits scans of business cards into individual images per card. It supports both PDF files (multi-page scans) and single image files. The output consists of cropped images for each card, named according to their page and grid position.

Usage

python scan_cards.py INPUT_FILE --nrows N --ncols M

• INPUT_FILE: Path to the input PDF or image file.
• --nrows N: Number of rows in the grid (N).
• --ncols M: Number of columns in the grid (M).

Example

python scan_cards.py cards.pdf --nrows 2 --ncols 3

This will split each page of cards.pdf into a 2x3 grid, saving each card as a separate PNG file.

Method of Operation

1. Input Handling: Accepts a PDF or image file. If PDF, each page is converted to an image.
2. Thresholding: Converts each image to grayscale and sets all pixels below a brightness of 20 to black.
3. Grid Splitting: Each image is divided into an NxM grid, cropping out each cell.
4. Saving Output: Each cell is saved as INPUT_pPAGENUM_imIMNUM.png.

Future Work

• Fine-tuning cell boundaries: Improve cropping to better fit card edges, possibly using edge detection or contour finding.
• Automatic grid detection: Detect grid layout automatically from the scan.
• Card content extraction: Integrate OCR or other methods to extract information from each card image.
• Batch processing: Support for processing multiple files or directories at once.

See requirements.txt for dependencies.