Toponym Recognition in Scanned Color Topographic Maps

Abstract

Topographic paper maps are a common support for geographical information. In the field of document analysis of this kind of support, this paper proposes an automatic approach to extract and recognize toponyms. We present a technique based on image segmentation and connected component processing. Different filtering stages ensure the consistency of plausible characters and strings. Detected text areas are used to feed an OCR software and the recognized words are analyzed and corrected. The main advantage of our technique is that no assumption is made about the character font, size or orientation. Experimental results obtained are encouraging in term of recognition efficiency.

Generating Named Road Vector Data from Raster Maps

“…Their orientation detection method only allows a string to be classified into 1 of the 4 directions. In both [Pouderoux et al, 2007;Roy et al, 2008], their methods do not hold when the string characters have very different heights or widths. Moreover, these approaches handle specific types of road labels and do not work further to determine the association between the recognized road labels and the geographic features in raster maps.…”

“…For text recognition from raster maps, Pouderoux et al [2007] present a text recognition technique for raster maps. They identify text strings in a map by analyzing the geometric properties of individual connected components in the map and then rotate the identified strings horizontally for OCR.…”

A Method for Automatically Extracting Road Layers from Raster Maps

“…Li et al [9] utilize a fixed color filter to extract the "black layer" from the USGS topographic maps and then work on the black layers to extract and rebuild the text labels and lines. Pouderoux et al [10] also use a preset grayscale threshold to generate a binary map for their toponym recognition algorithm, and the authors also note that additional methods such as the K-means color segmentation should be adopted to process more complex maps. Habib et al [7] work on the raster maps which contain only road lines to extract road intersections, and hence their binarization step is simply an edge detector.…”

Character Degradation Model and HMM Word Recognition System for Text Extracted from Maps

“…Due to the difficulties associated with the extraction of features from geographic maps in general, and topographic maps in particular, many of the systems presented in the literature have focused on simpler maps where features are well separated from each other or printed in colors that facilitate the map digitization process (e.g. see (Pouderoux et al, 2007), (Roy et al, 2007)). Other systems mostly consider the extraction of one or more types of features from complex maps ( (Chiang et al, 2005;Dhar & Chanda, 2006;Gamba & Mecocci, 1999;Kerle & Leeuw, 2009;Khotanzad & Zink, 2003;Leyk et al, 2006)), but do not address the problems of non-horizontal text or character recognition.…”