Researchers
Reginald Souleyrette
David Veneziano
About the research
This report evaluates the use of remotely sensed images in implementing the Iowa DOT LRS that is currently in the stages of system architecture. The Iowa Department of Transportation is investing a significant amount of time and resources into creation of a linear referencing system (LRS). A significant portion of the effort in implementing the system will be creation of a datum, which includes geographically locating anchor points and then measuring anchor section distances between those anchor points. Currently, system architecture and evaluation of different data collection methods to establish the LRS datum is being performed for the DOT by an outside consulting team.
This research extends that work by further evaluating the use of remotely sensed images for different components of the LRS. Specifically, the use of imagery for creation of the datum, including locating anchor points and measuring anchor sections; producing a spatial representation of the datum; and locating intermediate intersections along the datum were investigated. Three imagery datasets were evaluated in the various studies. They included a 6-inch resolution dataset, a 24-inch resolution dataset, and a 1-meter resolution dataset. The 1-meter dataset simulated the best satellite data available commercially. A 2-inch resolution dataset was evaluated for several of the studies. However, coverage in the images was limited so it could not be fully evaluated. Additionally, Roadware consultants agreed and measured five test segments (located in Pilot Study Area 2) using a DMI and DGPS as part of pavement condition assessment they were conducting.
The first section of this report provides background information on remote sensing. Section 3 describes the datasets and pilot study areas that were included to complete the research. The fourth section discusses the use of imagery to establish the geographic locations of anchor points and business data. The actual spatial accuracy of the images is evaluated as well as how well features can actually be identified at different levels of image resolution. Additionally, the human error that may result due to variation in the manner that observers manually locate objects in images was evaluated and reported for the four image datasets. Section 5 investigated the accuracy with which the imagery and Roadware methods could measure anchor section distances as compared to the VideoLog DMI data that were collected as part of the Iowa DOT LRS Pilot Study. Next, the use of imagery and DGPS was evaluated for use in creation of a spatial representation of the datum. Finally, different methods were compared for calculation of the distance along anchor sections to intermediate, non-anchor point intersections. The imagery datasets and the use of GIMS cartography were discussed.