Text Extraction from Historical Handwritten Documents by Edge Detection

Many national archives or libraries keep large amount of historical handwritten documents. One problem that many archivists are facing is the sipping of ink through the pages of certain double-sided handwritten documents after long periods of storage. The result is that the handwritten characters from the reverse side appear as noise on the front side and even interfere with the front side characters, which makes the reading of the contents quite difficult. This paper addresses this problem and develops a novel method to extract clear textual images from interfering and overlapping areas. With the critical observation that the edges of the sipping strokes from the reverse side are not as sharp as those on the front side, we adopt an edge detection approach to suppress unwanted background patterns. Other remaining long and strong noisy edges are further removed by using an orientation filter and a size filter. The proposed method proves to perform well regardless of the intensity differences between the image on the front side and the interfering image from the reverse side. The segmentation results of real images are shown and evaluated.