Text line extraction from multi-skewed handwritten documents

A text line extraction method is presented for multi-skewed handwritten document images. First the whole document image is divided into several sub-images with overlapped area, then for every sub-image, it assumes that hypothetical water flows, from both left and right sides of the image frame, face obstruction from characters of text lines. The stripes of areas left unwetted on the sub-images are labeled for extraction of text lines. Finally, touching lines are segmented by detecting the bottom edge points. The experimental results show the approach is effective and available to extract curved text lines.