Efficient and Robust Graphics Recognition from Historical Maps

Abstract

Abstract. Historical maps contain rich cartographic information, such as road networks, but this information is "locked" in images and inaccessible to a geographic information system (GIS). Manual map digitization requires intensive user effort and cannot handle a large number of maps. Previous approaches for automatic map processing generally require expert knowledge in order to fine-tune parameters of the applied graphics recognition techniques and thus are not readily usable for non-expert users. This paper presents an efficient and effective graphics recognition technique that employs interactive user intervention procedures for processing historical raster maps with limited graphical quality. The interactive procedures are performed on color-segmented preprocessing results and are based on straightforward user training processes, which minimize the required user effort for map digitization. This graphics recognition technique eliminates the need for expert users in digitizing map images and provides opportunities to derive unique data for spatiotemporal research by facilitating timeconsuming map digitization efforts. The described technique generated accurate road vector data from a historical map image and reduced the time for manual map digitization by 38%. Keywords:Color image segmentation, road vectorization, historical raster maps, image cleaning IntroductionMaps contain valuable cartographic information, such as locations of historical places, contour lines, building footprints, and hydrography. Extracting such cartographic information from maps (i.e., creating spatial layers that can be processed in a GIS) would support multiple applications and research fields. For example, there are numerous cases in which historical maps have been used to carry out research in land-cover change and biogeography [Kozak et al., 2007;Petit and Lambin, 2002], and urban-area development [Dietzel et al., 2005].Today, thousands of such maps and map series are available in scanned raster format (i.e., digital map images) in a variety of digital archives. Previous work on extracting cartographic information from raster maps typically requires intensive user intervention for training and parameter tuning, in particular, when processing historical maps of poor graphical quality [Gamba and Mecocci, 1999;Leyk and Boesch, 2010].

Automatic Extraction of Forests from Historical Maps Based on Unsupervised Classification in the CIELab Color Space

“…The proposed approach provides a general *1900) framework to deal with this classical issue. Chiang et al (2012) also adopted morphological filters in a pre-processing step. He applied successive erosions to remove elevation contour lines before to extract the roads which are very similar.…”

“…If the maps contain thematic colors layers, color signature can provide much information to capture objects of interest but means some limitations when processing poor image quality (Leyk 2006). Chiang et al (2012) used K-means algorithm to capture road vectors in a raster map. Results show that algorithm is efficient but that the users need to determinate a large number of K clusters to separate different features.…”

An Adaptable Approach for Generating Vector Features from Scanned Historical Thematic Maps Using Image Enhancement and Remote Sensing Techniques in a Geographic Information System

“…Segmenting the image based on color has been utilized effectively to separate features in scanned historical maps (Leyk 2010;Leyk and Boesch 2010). Even further, Chiang, Leyk, and Knoblock (2013) investigated graphics recognition techniques that required interactive user participation, while Baily et al (2011), Dhar, Bikash, andChanda (2006), and Oka, Garg, and Varghese (2012) detailed automatic vectorization of features from maps using varying techniques and methodologies. One of the most recent approaches to solving the raster-to-vector conversion problem came when Herrault et al (2013) utilized an unsupervised classification approach to achieve automatic extraction of forests from historical maps.…”

Automatic Text Removal and Replacement in Scanned Meteorological Thematic Maps

“…With the increasingly popular computing-on-the-go devices, detecting texts of arbitrary orientations for images captured by devices such as camera phones and tablets that are under less ordered conditions has become increasingly important and at the same time is a challenging task. Chiang et al (2013) proposed an algorithm to detect text of random orientations in typical images. The algorithm is based on a two-level classification arrangement using two sets of features and is specially designed for capturing both essential and orientation-invariant characteristics of text.…”