Ching-Chien Chen

Automatic and Accurate Extraction of Road Intersections from Raster Maps

Since maps are widely available for many areas around the globe, they provide a valuable resource to help understand other geospatial sources such as to identify roads or to annotate buildings in imagery. To utilize the maps for understanding other geospatial sources, one of the most valuable types of information we need from the map is the road network, because the roads are common features used across different geospatial data sets. Specifically, the set of road intersections of the map provides key information about the road network, which includes the location of the road junctions, the number of roads that meet at the intersections (i.e., connectivity), and the orientations of these roads. The set of road intersections helps to identify roads on imagery by serving as initial seed templates to locate road pixels. Moreover, a conflation system can use the road intersections as reference features (i.e., control point set) to align the map with other geospatial sources, such as aerial imagery or vector data. In this paper, we present a framework for automatically and accurately extracting road intersections from raster maps. Identifying the road intersections is difficult because raster maps typically contain much information such as roads, symbols, characters, or even contour lines. We combine a variety of image processing and graphics recognition methods to automatically separate roads from the raster map and then extract the road intersections. The extracted information includes a set of road intersection positions, the road connectivity, and road orientations. For the problem of road intersection extraction, our approach achieves over 95% precision (correctness) with over 75% recall (completeness) on average on a set of 70 raster maps from a variety of sources.

Automatic extraction of road intersections from raster maps

Numerous raster maps are available on the Internet, but the geographic coordinates of the maps are often unknown. In order to determine the precise location of a raster map, we exploit the fact that the layout of the road intersections within a certain area can be used to determine the maps location. In this paper, we describe an approach to automatically extract road intersections from arbitrary raster maps. Identifying the road intersections is difficult because raster maps typically contain multiple layers that represent roads, buildings, symbols, street names, or even contour lines, and the road layer needs to be automatically separated from other layers before road intersections can be extracted. We combine a variety of image processing and graphics recognition methods to automatically eliminate the other layers and then extract the road intersection points. During the extraction process, we determine the intersection connectivity (i.e., number of roads that meet at an intersection) and the road orientations. This information helps in matching the extracted intersections with intersections from known sources (e.g., vector data or satellite imagery). For the problem of road intersection extraction, we applied the techniques to a set of 48 randomly selected raster maps from various sources and achieved over 90% precision with over 75% recall. These results are sufficient to automatically align raster maps with other geographic sources, which makes it possible to determine the precise coverage and scale of the raster maps.

Automatically Annotating and Integrating Spatial Datasets

Recent growth of the geo-spatial information on the web has made it possible to easily access a wide variety of spatial data. By integrating these spatial datasets, one can support a rich set of queries that could not have been answered given any of these sets in isolation. However, accurately integrating geo-spatial data from different data sources is a challenging task. This is because spatial data obtained from various data sources may have different projections, different accuracy levels and different formats (e.g. raster or vector format). In this paper, we describe an information integration approach, which utilizes various geo-spatial and textual data available on the Internet to automatically annotate and conflate satellite imagery with vector datasets. We describe two techniques to automatically generate control point pairs from the satellite imagery and vector data to perform the conflation. The first technique generates the control point pairs by integrating information from different online sources. The second technique exploits the information from the vector data to perform localized image-processing on the satellite imagery. Using these techniques, we can automatically integrate vector data with satellite imagery or align multiple satellite images of the same area. Our automatic conflation techniques can automatically identify the roads in satellite imagery with an average error of 8.61 meters compared to the original error of 26.19 meters for the city of El Segundo and 7.48 meters compared to 15.27 meters for the city of Adams Morgan in Washington, DC.

A general approach to discovering, registering,and extracting features from raster maps

Maps can be a great source of information for a given geographic region, but they can be difficult to find and even harder to process. A significant problem is that many interesting and useful maps are only available in raster format, and even worse many maps have been poorly scanned and they are often compressed with lossy compression algorithms. Furthermore, for many of these maps there is no meta data providing the geographic coordinates, scale, or projection. Previous research on map processing has developed techniques that typically work on maps from a single map source. In contrast, we have developed a general approach to finding and processing street maps. This includes techniques for discovering maps online, extracting geographic and textual features from maps, using the extracted features to determine the geographic coordinates of the maps, and aligning the maps with imagery. The resulting system can find, register, and extract a variety of features from raster maps, which can then be used for various applications, such as annotating satellite imagery, creating and updating maps, or constructing detailed gazetteers.

Automatic alignment of vector data and orthoimagery for the national map

A general problem in combining road vector data with orthoimagery from different sources is that they rarely align. There are a variety of causes to this problem, but the most common one is that the latest products are collected with higher accuracy and improved processing techniques. In previous work, we developed techniques to automatically correct the alignment of vector data with orthoimagery using a technique called conflation. However, in applying our technique to real-world datasets provided by USGS, we discovered that these techniques failed in some areas. In this paper, we describe some refinements to our original approach that provide consistently better results in aligning the vector data with the orthoimagery.