Convolutional Neural Networks for Page Segmentation of Historical Document Images
This addresses document digitization for archivists and historians, but it is incremental as it applies a known CNN approach to a specific domain with a simplified architecture.
The paper tackles page segmentation of historical document images by framing it as a pixel labeling problem and using a simple CNN with one convolution layer, achieving competitive results on public datasets.
This paper presents a Convolutional Neural Network (CNN) based page segmentation method for handwritten historical document images. We consider page segmentation as a pixel labeling problem, i.e., each pixel is classified as one of the predefined classes. Traditional methods in this area rely on carefully hand-crafted features or large amounts of prior knowledge. In contrast, we propose to learn features from raw image pixels using a CNN. While many researchers focus on developing deep CNN architectures to solve different problems, we train a simple CNN with only one convolution layer. We show that the simple architecture achieves competitive results against other deep architectures on different public datasets. Experiments also demonstrate the effectiveness and superiority of the proposed method compared to previous methods.