Class Quantification of Aerial Images using Maximum Likelihood Estimation

Class quantification of aerial images plays a vital role in remote sensing. One of the class quantification method is discussed in this paper. Proposed method uses Maximum Likelihood Estimation based classifier for class quantification. Algorithm is trained by the sample classes derived from parent image. Feature space is estimated from each training sample. Different classes are labeled in test image by maximizing the likelihood function. The experimentation is done on aerial images obtained by Geo eye satellite at the elevation of 0.6km. The percentage area covered by the labeled classes is computed for all test images.