Interval Type-2 Fuzzy Based Neural Network For High Resolution Remote Sensing Image Segmentation

This paper presented a supervised image segmentation approach based on interval type-2 fuzzy neural network to overcome the problems brought by high resolution remote sensing images. The interval type-2 model was obtained through blurring the mean and variance of the Gaussian model which characterized the uncertainly of the membership of pixels. Then the fuzzy membership and its upper and lower fuzzy membership of the training samples were used as the input of the neuron network in which the influence of neighbor pixels were taken into consideration to construct a decision model to realize the segmentation. The proposed algorithm, FCM and HMRF-FCM algorithm and an interval type-2 fuzzy neuron networks without spatial relationships were performed on synthetic and real high resolution remote sensing images. The qualitative and quantitative analyses demonstrate the efficient of the proposed algorithm, especially for homogeneous regions which contains a great difference in its gray level.