Object-based Land Cover Classification and Change Analysis in the Baltimore Metropolitan Area Using Multitemporal High Resolution Remote Sensing Data

Abstract: Accurate and timely information about land cover pattern and change in urbanareas is crucial for urban land management decision-making, ecosystem monitoring andurban planning. This paper presents the methods and results of an object-basedclassification and post-classification change detection of multitemporal high-spatialresolution Emerge aerial imagery in the Gwynns Falls watershed from 1999 to 2004. TheGwynns Falls watershed includes portions of Baltimore City and Baltimore County,Maryland, USA. An object-based approach was first applied to implement the land coverclassification separately for each of the two years. The overall accuracies of theclassification maps of 1999 and 2004 were 92.3% and 93.7%, respectively. Following theclassification, we conducted a comparison of two different land cover change detectionmethods: traditional (i.e., pixel-based) post-classification comparison and object-basedpost-classification comparison. The results from our analyses indicated that an objectbasedapproach provides a better means for change detection than a pixel based methodbecause it provides an effective way to incorporate spatial information and expertknowledge into the change detection process. The overall accuracy of the change mapproduced by the object-based method was 90.0%, with Kappa statistic of 0.854, whereasthe overall accuracy and Kappa statistic of that by the pixel-based method were 81.3% and0.712, respectively.