This is a refactored version of the document scanner I wrote for my Computational Photography class. It uses an edged image, computes its Hough Transform, finds the best 4 lines, computes the corners, and rectifies the image by applying a Homography Transformation.
The following dependencies are required:
- Computer Vision Toolbox
The high level files are main.m and scanner.m. The main script allows for picking the image you wish to "scan", and scanner calls all the lower level functions to compute the rectified image.
Kevin Karnani
Adin Solomon