Search

This paper proposes an effective multiscale method for the segmentation of the synthetic aperture radar (SAR) images via probabilistic neural network. By combining the probabilistic neural network (PNN) with the multiscale autoregressive (MAR) model, a classifier, which inherits the excellent strongpoint from both of them, is designed. The MAR models are utilized to extract the multiscale feature of SAR image, which is the input of the network. The PNN is trained by the proposed algorithm, and then the SAR images are segmented by the trained network. The experimental result demonstrates the effectiveness and efficiency of the proposed method.

A novel scheme of digital image watermarking based on the combination of dual-tree wavelet transform (DTCWT) and probabilistic neural network is proposed in this paper. Firstly, the original image is decomposed by DTCWT, and then the watermark bits are added to the selected coefficients blocks. Because of the learning and adaptive capabilities of neural networks, the trained neural networks can recover the watermark from the watermarked images. Experimental results show that the proposed scheme has good performance against several attacks.

A method for the segmentation of synthetic aperture radar (SAR) image is presented in this paper. The method integrates the use of multi-scale technology, mixed-model information and support vector machines (SVM). First, the multi-scale autoregressive (MAR) model is modeled for multi-scale sequence of SAR image, and a multi-scale features, which is used as input of SVM, are extracted via the MAR model. Then, SVM is trained and the SAR image is segmented with the trained SVM. So, this method not only can be fully taken advantage of the statistical information of SAR images in multi-scale sequence but also ability of SVM classifier. The experimental results show that the method has a very effective computational behavior and effectiveness, and decrease the time and increase the quality of SAR image segmentation.

A new digital watermarking algorithm based on BPN neural network is proposed. Watermark embed processing is carried out by transforming the host image in wavelet domain. Watermark bits are added to the selected coefficients blocks. Because of the learning and adaptive capabilities of neural networks, the trained neural networks can recover the watermark from the watermarked images. The experimental results show that this watermarking algorithm has a good preferment.