An implementation of some binarization methods such as Niblack, Sauvola, Wolfjolion [1] and one based on feature space partitioning that uses the others as auxiliary methods [2].
Compile using the Makefile. There is also a Shell Script that makes it possible to run the code with different input images and different binarization methods.
This code is ugly and was made in a hurry. But works pretty well. One of the major needed refactorings is use just one version of OpenCV, and make better usage of space/memory.
Check directories input and output for input images and output images using different binarization methods.
For example, for this input image (http://homes.dcc.ufba.br/~caiosba/mestrado/binarization/d.png), this one (http://homes.dcc.ufba.br/~caiosba/mestrado/binarization/d.this.w.png) is produced as output when using the feature space partitioning algorithm with Wolfjolion as auxiliary method.
The program can be run this way:
./binary <input image name without extension (it will assume .png)> <binarization method>
Binarization method can be one of the following: s (Sauvola), n (Niblack) or w (Wolfjolion). Each execution will produce two images: one using the method as primary method and another using it as an auxliary method for the feature space partitioning binarization.
Developed as a work for the master lecture Topics on Visual Computing, by Profs. Vinícius and Perfilino, at Federal University of Bahia - Brazil.
References: