Neighborhood Correlation Image Analysis Technique for Change Detection in Forest Landscape

The origin of the concept of using contextual information of Correlation in neighborhood of a pixel for change detection is the simple geo-statistical fact that the same geographical area (neighborhood window) on two dates of imagery will tend to be highly correlated if no or little change has occurred and uncorrelated when change occurs. The main objective of this study is to investigate the usability of this technique for change detection in forest landscape with the use of medium spatial resolution satellite data with decision tree classification approach. This research has came to the conclusion that the NCI analysis technique does provide change information and it requires larger windows (15 x 15 pixels) for NCI information to become useful when medium resolution imagery is used for change detection in forest.