Comparison Raster data and Rasterized Vector Data with ‘Saw Tooth Line Algorithm’ for Verifying the Accuracy of Digital Map

Songhoon Baik
Member of Technical Staff, GIS Service Development Team, Outside Plant Technology Laboratory, Korea Telecom, Republic of Korea

It has been essential to develop a proper methodology for quality assurance in digital map (or vector map, hereinafter referred to as digital map) producing process since GIS has been recognized as useful in utilities, national agencies and municipalities and so on.

The purpose of this study is to develop computer-based quality assurance tool which can be adapted in producing digital map. Usually, there are three ways to produce digital map. Those are manual digitizing, screen digitizing, and mapping with analytical stereo plotter. In Korea, screen digitizing is a prevailing method in producing digital map.

Considering the above mentioned conditon, the devoloped tool is intended for the digital map with screen digitized through scanning paper map.

Getting accuracy of base map database is one of the most important factors to ensure successful Korea’s National GIS construction. The conventional methods for quality assurance, however, have depended on inaccurate and ‘time-consuning’ manual works which may cause potential problems in national wide GIS deployment.

This study has started to develop a software tools which can be used to verify accuracy of digitized maps automatically and effectively.

Inaccuracy and errors may occurred either in every step from scanning paper map through digitizing, or may caused by individual workers’ mistakes. Considering all anticipated inaccuracy and errors throughly for each step in the whole process, "Similiarity Measuring Theory" was introduced to the system development. By using ‘Brensenham Algorithm’ and ‘Saw Tooth Line Algorithm’, digitized vector maps are converted into raster maps automatically and overlapped over the original scanned raster maps, which is used for the source of vectorization for the verification of accuracy. Completing the verification, the system automatically report the differences between the original scanned raster data maps and the rasterized vector maps.

Figure1 shows the examples of raster data map (a) through scanning the original paper map, and the rasterized vector data map (b). The symbol O marks the differences between the two types of data map. O are the errors which are missed in process of digitizing the raster data map. As shown, raster data map has some noises. But these are not errors. Therefore these noises are eliminated by filtering.

Figure 1: Raster data map with scanning paper map (a), and rasterized vector data map (b)

Figure 2: XOR operating (a), and filtering (b) over two types of data map

The system also provides the method for positional accuracy measurement of point features in digitized maps. A digitized vector map, which is a DXF (Drawing eXchange Format) file, in TM coordinate system has control points which will be compared with corresponding points on the original paper map.

This verification tool has been reflected by the examples of Ordnance Survey and Nation Joint Utilites Group.

This tool will be expected to replace the conventional verifying method which depends on inspector’s manual work, and to save the man-hour and cost. In addition, this tool will be expected to make great contribution not only to GIS projects in Korea Telecom including

TOMS (Telephone Outside-plant Management System) but also to National GIS project which is being propelled by government.