| Abstract: |
While there is an abundance of land cover data available nationwide, such data is often too coarse to be meaningful for local decision making. Cities and towns are often rich in remotely sensed data thanks to robust orthophotography programs and the availability of high-resolution commercial satellite imagery, yet they lack the corresponding detailed land cover information. Manual interpretation of high-resolution remotely sensed data is too time consuming and costly while traditional "pixel based" classifiers perform poorly when applied to high-resolution imagery. Over the past few years there have been considerable advances in made in automating the process of deriving detailed land cover information from high-resolution remotely sensed data. Using examples from Burlington, Baltimore, and Washington DC this presentation will show how automated feature extraction technology is being employed to create accurate, GIS-ready, high-resolution land cover datasets. |