I have implemented this following paper here :
Jia, F., Shi, C., He, K., Wang, C., & Xiao, B. (2016).
Document image binarization using structural symmetry of strokes. Frontiers in Handwriting Recognition (ICFHR), 2016 15th International Conference On, 411–416.
Although some part of this implementation is taken from the journal version of this conference paper :
Jia, F., Shi, C., He, K., Wang, C., & Xiao, B. (2018).
Degraded document image binarization using structural symmetry of strokes. Pattern Recognition, 74, 225–240.
The following parts are taken from the journal version of this paper (which is supposed to be more robust and better performing)
1) Background image estimation (actually taken from this paper :
Ntirogiannis, K., Gatos, B., & Pratikakis, I. (2014).
A combined approach for the binarization of handwritten document images. Pattern Recognition Letters, 35(1), 3–15.)
-
Image normalization
-
Gradient image calculation and pixel direction calculation from gradient image
Then the remaining portion are taken from the above mentioned paper :
Jia, F., Shi, C., He, K., Wang, C., & Xiao, B. (2016).
Document image binarization using structural symmetry of strokes. Frontiers in Handwriting Recognition (ICFHR), 2016 15th International Conference On, 411–416.
The main function to call :
Structural_Symmetric_Strokes.m