Document image binarization result of manuscript "Binarization of Unevenly Illuminated Document Images Based on Cloth Simulation Filter"

Document image binarization is a crucial step in OCR algorithms. A binarization method that is robust to unevenly illuminated document images can provide better inputs for algorithms to recognize text under complex lighting conditions. By considering document images with uneven illumination as a terrain point cloud in 3D space, we transformed the problem of handling an uneven lighting background into a problem of ground filtering in the field of point cloud processing. We propose a binarization method based on the cloth simulation filter (CSF), a point cloud processing method, for binarizing unevenly illuminated document images. This dataset contains binarization results in the manuscript "Binarization of Unevenly Illuminated Document Images Based on Cloth Simulation Filter". The comparing methods include Adaptive, Bataineh, CSF(proposed), ISauvola, Sauvola, Su, TRSingh, and Wolf. All were applied with 176 images from the WEZUT dataset.文档图像二值化是光学字符识别（Optical Character Recognition，OCR）算法中的关键步骤。对光照不均的文档图像具备鲁棒性的二值化方法，可为复杂光照条件下的文本识别算法提供更优质的输入数据。我们将光照不均的文档图像视作三维空间中的地形点云，将处理不均匀光照背景的问题转化为点云处理领域内的地面滤波问题。我们提出了一种基于点云处理方法布料模拟滤波器（Cloth Simulation Filter，CSF）的二值化方法，用于实现光照不均文档图像的二值化处理。本数据集包含论文《基于布料模拟滤波器的光照不均文档图像二值化》中的二值化结果。本次对比的方法包括自适应二值法（Adaptive）、Bataineh法、所提CSF法、ISauvola法、Sauvola法、Su法、TRSingh法以及Wolf法。所有方法均基于WEZUT数据集中的176张图像开展测试。