ZIBA: A Novel Algorithm Based on Zero-Sum Game For Document Image Binarization
Pouria Khanzadi, Mehrnoush Vaseghipanah, Shirin Kordnoori, Hamidreza Mostafaei · 2019
In this paper, we proposed an algorithm based on game theory for document image binarization. For designing the algorithm, we formulate document image binarization using zero-sum game in which foreground of the image with textual shapes and background of the image are considered as two players. Then the best distance between the players is calculated. Finally, a global threshold is set for binarization of the image. The proposed algorithm is called Zero-sum Image Binarization Algorithm (ZIBA). The experimental results show that ZIBA has admissible precision for document image binarization.