A review of recent advances in scanned topographic map processing

Historical topographic maps carry huge information of this planet over a very long period. Geographic information extraction from topographic map (TM) is helpful for understanding the environmental changes and other geography related research fields. However, both quality and quantity of TM bring big challenges in efficient interpretation. Numerous studies have been focused on extracting geographic information from scanned topographic maps (STM). In this work, we will review lots of studies on STM processing in the past a few decades. Most of the discussing approaches are grouped into two categories which are segmentation and vectorization. Segmentation is the foundation for information extraction from STM and the segmenting results directly influence the subsequent procedures. Vectorization component contains four main geographic elements including contour line, road, text and symbols. This work reviews amount of published articles in these research areas and gives brief introduces about those previous studies as well as analysis and comments. Our goal is to show the exist ideas in STM processing and provide some general knowledge on methods and patterns which are employed in those studies. Most important, through the overview of existing approaches, new thoughts and ideas about STM processing are expected. (c) 2018 Elsevier B.V. All rights reserved.