A Combined Layer-Based Approach for the Segmentation of Document Images

摘要

There are many challenges addressed in MRC compression, such as the segmentation of the document into foreground and background layers, represented as a binary mask. Usually, the resulting quality of a MRC compression method depends on the segmentation algorithm used to compute the binary mask. The aim of this work is to leverage state-of-the-art techniques in document segmentation allowing further content, such as text or images, to be exploited in an efficient way. Then, a novel methodology is proposed that leads to increasing document segmentation performance for MRC compression scheme.