SPIE

A new method for classifying land cover features in high resolution satellite imagery was developed. The method involves (1) performing image segmentation to delineate homogeneous ground objects in the image, and (2) using a convolutional neural network classifier to extract the land cover of the segments (i.e. whether they represent buildings...

Multiresolution segmentation and rule-based classification techniques are used to classify objects from very high-resolution satellite images of urban areas. The major contribution of this research is the development of rule sets and the identification of major features for satellite image classification where the rule sets are transferable and the...