Recently, neural network has been increasingly applied to remote sensing imagery classification. The conventional neural network classifier performs learning from the representative information within a problem domain on a one-pixel-one-class basis; therefore, class mixture and the degree of membership of a pixel are generally not taken into account, often resulting in a poor classification accuracy. Based on the framework of a dynamic learning neural network (DL), this communications proposes a fuzzy version (FDL) based on two steps: network representation of fuzzy logic and assignment of membership. Comparisons between the DL and FDL are made by applying both neural networks to SAR image classification. Experimental results show that the FDL has faster convergence rate than that of DL. In addition, the separability between similar classes is improved. Moreover, the classification results match better with ground truth.Index Terms-Fuzzy logic, neural network.

No abstract

In this paper, a new classification scheme of polarimetric synthetic aperture radar (PolSAR) images using fractal dimension as contextual information is proposed. Support vector machines (SVM) due to their ability to handle the nonlinear classifier problem are applied to a new fractal feature vector, which is constructed from Pauli decomposed vector and fractal dimensions. Fractal dimension is computed based on the concepts of fractional Brownian motion (fBm) and wavelet multi-resolution analysis using a self-adaptive window approach and fuzzy logic. The experimental results on AIRSAR images prove effectiveness of the proposed vector in comparison to the Pauli decomposed vector.

The neural network learning process is to adjust the network weights to the selected training data. Based on the multi-layer feed-forward pcrceptron neural network, a dynamic learning algorithm (DL) is proposed in this paper. The presented learning algorithm makes use of Kalman filtering technique to update the network weights, in the sense that the stochastic characteristics of incoming data sets are implicitly incorporated into the network. The Kalman gains which represent the learning rates of the network weights are updated and calculated through the U-D factorization.By concatenating all of the network weights at each layer to form a long vector such that it can be updated without propagating back. the proposed algorithm improves in convergence substantially ovcr the back-propagation (BP) learning algorithm. Numerical illustrations are carried out using two types of problems: multispectral imagery classification and surface parameter retrieval. R C S U~L~ indicate that the usc of Kalman filtering algorithm not only substantially improves the convergence rate in the learning stage. but also enhances the separability for prohlems with highly nonlinear boundaries, as compared to BP algorithm. suggesting that the proposed DL neural network provides a practical and efficient tool for remotc sensing applications.

No abstract