<![CDATA[This paper presents the result of continuing study on image classification. In previous study, have recommended a neuro-satistical scheme in the framework of multitemporal optical sensor image classification. Thr scheme consists of neural network classifier to compute the posterior probabilities, expectation maximum method to optimize prior join probabilities, compound probabilities to produce thematic image and change image. This paper report the result of extending the scheme for multidate multisensor image classification. For each sensor image classifier, two scheme have been evaluated. The first scheme has use the co-occurrence matrix  feature images or original tonal images as the input data and the gaussian kernel for the neural network classifier. The second scheme has use the original tonal image as the input data and the multinomial co-occurrence matrix kernel for the neural network classifier. The result are also compared to use of bachh propagation neural network classifier. Base on this study we have proposed a scheme for multidate multisensor image classifier.t]]>
Copyrights © 2011
