Combining Morphological and Histogram based Text Line Segmentation in the OCR Context

Text line segmentation is one of the pre-stages of modern optical characterrecognition systems. The algorithmic approach proposed by this paper has beendesigned for this exact purpose. Its main characteristic is the combination oftwo different techniques, morphological image operations and horizontalhistogram projections. The method was developed to be applied on a historicdata collection that commonly features quality issues, such as degraded paper,blurred text, or presence of noise. For that reason, the segmenter in questioncould be of particular interest for cultural institutions, that want access torobust line bounding boxes for a given historic document. Because of thepromising segmentation results that are joined by low computational cost, thealgorithm was incorporated into the OCR pipeline of the National Library ofLuxembourg, in the context of the initiative of reprocessing their historicnewspaper collection. The general contribution of this paper is to outline theapproach and to evaluate the gains in terms of accuracy and speed, comparing itto the segmentation algorithm bundled with the used open source OCR software.Comment: Journal of Data Mining and Digital Humanities; Small adjustment