EMGU CV

Emgu cv is a wrapper for OpenCV in C #.

It contains almost all features of OpenCV to perform all basic tasks of image analysis and processing. There are also tools of machine learning such as SVM, Neural Networks, Naive Bayes, Decision Tree …

Since version 2.3, there is also the Tesseract OCR included.

Examples are available with the libraries as : face detection, SVM, motion detection, pedestrian detection, recognition panels, SURF, number plate recognition.

http://www.emgu.com/wiki/index.php/Main_Page

 

Voronoi

The Voronoi diagram is commonly used to decompose a space from points or set of points.

If we define a space containing n points, for each point p, the Voronoi cell can be drawn. It defines the set of points in space that are closer to p than to any other point.

 

Koichi Kise, Akinori Sato and Motoi Iwata propose a method for document segmentation based on the Voronoi diagram. The principle consists to discretize the surface contours and then to apply Voronoi to groups of points.

Reference:
K. Kise, A. Sato, and M. Iwata. Segmentation of page images using the area Voronoi diagram. Computer Vision and Image Un- derstanding, 70(3) :370{382, 1998.

RXY cut

The recursive XY cut (RXY cut) is a top down technique used for page segmentation. The algorithm recursively split the original image into sub-rectangles. To do this, the vertical and horizontal profiles are plotted (they correspond to the sum of pixels along the X axis and along the Y axis). The split is done recursively on the most dense area space.

Creation of horizontal projection profil
Examples of layout that cannot be segmented by RXY cut.

Reference :
G. Nagy and S. Seth. Hierarchical representation of optically scanned documents. In Proc. of the 17th Conf. on Pattern Recognition, pp. 347–349, 1984.

RLSA

The Run Length Smoothing Algorithm WAS published in 1981 by L. Abele and F. Wahl for document segmentation. Overall, the idea is to connect the black pixels separated by less than n pixels white. If n is increased, we can segment letters, words, lines and paragraphs.

The RLSA is usually applied horizontally and vertically. This two operations are then combined with a logical AND in order to obtain the final result.

rlsa
rlsa2

Reference:
K.Y. Wong, R.G. Casey and F.M. Wahl, “Document analysis system,” IBM J. Res. Devel., Vol. 26, NO. 6,111). 647-656, 1982.