Technology Profile: Between the Lines
The State of Maryland has hundreds of thousands of road features that are distributed along a transportation network more than 5,000 miles long. In early 2009, the Maryland State Highway Administration (SHA) faced a critical need to identify all painted roadway markings and create an accurate inventory to help ensure that the painted lines were being properly maintained.
Traditional field-based methods to find highway markings were prohibitively expensive and time consuming. “Highway features have to be maintained and kept up to date, and it is not cost effective to track highway features by driving through the state and manually tracking them,” explains Brad Davis, a former GIS liaison with the Maryland SHA Planning Department. What’s more, the information often becomes obsolete at the same rate it’s being collected.
To curb costs and to have a more accurate and timely inventory, the Maryland SHA elected to use the new ENVI aerial imagery and image analysis software from ITT.
The use of satellite and aerial imagery is rapidly gaining popularity for a variety of applications as it becomes less expensive and more readily available. Such imagery can help provide information about a geographic area of interest from an overall or detailed perspective and can be used to update a GIS for mapping purposes.
Using aerial imagery obtained by the State of Maryland for other uses, Davis was tasked with applying the ENVI software to the inventory of painted roadway lines across Maryland. One of the advanced capabilities in ENVI is the feature extraction tool, which allows users to find and select features in an image. The software automatically segments the image into regions of pixels, computes attributes for each region to create objects, and classifies the objects (with rule-based or supervised classification) based on those attributes to extract the features. This automated workflow allowed Davis to identify white and yellow lines painted on the road surfaces with the ease and speed needed to create an accurate inventory and update the inventory as needed in a timely manner.
Davis was also able to produce more results by identifying multiple types of painted lines. “By using ENVI’s feature extraction workflow, I was able to identify a large amount of dashed white lines and separate them by type, resulting in a more accurate and complete inventory,” Davis explains. In the end, Davis was able to discern between white-dashed and solid-line markings, yellow double- and single-line markings, and left-turn-only line markings.
Once lines were extracted, he created an inventory based on that information and exported the inventory to ESRI’s ArcGIS software. He then combined the processed imagery with road centerline information in the GIS. The final product was sent to a geodatabase, where it can be accessed anytime by Maryland SHA asset managers and engineers.
By reorienting the methods used to manage the highway assets, applying new techniques to aerial imagery of the Maryland transportation network, and implementing an automated method to extract the location and condition of painted roadway lines, the Maryland SHA has been able to stem the costs and time required for manual maintenance inventory and has significantly increased the accuracy of its inventory.
Other applications for aerial and satellite imagery and image processing and analysis tools are on the horizon. Once imagery is available, advanced software can provide a wealth of information, including road shoulder widths, guardrail and sign inventory, and much more. “Ultimately it’s another tool to make our processes more efficient,” Davis says.
For more information about ENVI, visit www.ittvis.com.