This paper proposes using a backpropagation (BP) neural network for the classification of ship targets in airborne synthetic aperture radar (SAR) imagery. The ship targets consisted of 2 destroyers, 2 cruisers, 2 aircraft carriers a frigate and a supply ship. A SAR image simulator was employed to generate a training set, a validation set, and a test set for the BP classifier.The features required for classification were extracted from the SAR imagery using three different methods. The first method used a reduced resolution version of the whole SAR image as input to the BP classifier using simple averaging. The other two methods used the SAR image range profile either before or after a local-statistics noise filtering algorithm for speckle reduction. Performance on an extensive test set demonstrated the performance and computational advantages of applying the neural classification approach to targets in airborne SAR imagery. Improvements due to the use of multi-resolution features were also observed.

A recent neural clustering scheme called "probabilistic winner-take-all (PWTA)" is applied to image segmentation. It is demonstrated that PWTA avoids underutilization of clusters by adapting the form of the cluster-conditional probability density function as clustering proceeds. A modification to PWTA is introduced so as to explicitly utilize the spatial continuity of image regions and thus improve the PWTA segmentation performance. The effectiveness of PWTA is then demonstrated through the segmentation of airborne synthetic aperture radar (SAR) images of ocean surfaces so as to detect ship signatures, where an approach is proposed to find a suitable value for the number of clusters required for this application. Results show that PWTA gives high segmentation quality and significantly outperforms four other segmentation techniques, namely, 1) K-means, 2) maximum likelihood (ML), 3) backpropagation network (BPN), and 4) histogram thresholding.

No abstract

Recent research has linked backpropagation (BP) and radial basis function (RBF) network classifiers, trained by minimizing the standard mean square error (MSE), to two main topics in statistical pattern recognition (SPR), namely the Bayes decision theory and discriminant analysis. However, so far, the establishment of these links has resulted in only a few practical applications for training, using, and evaluating these classifiers. The paper aims at providing more of these applications. It first illustrates that while training a linear output BP network, the explicit utilization of the network discriminant capability leads to an improvement in its classification performance. Then, for linear output BP and RBF networks, the paper defines a new generalization measure that provides information about the closeness of the network classification performance to the optimal performance. The estimation procedure of this measure is described and its use as an efficient criterion for terminating the learning algorithm and choosing the network topology is explained. The paper finally proposes an upper bound on the number of hidden units needed by an RBF network classifier to achieve an arbitrary value of the minimized MSE. Experimental results are presented to validate all proposed applications.

This paper presents the implementation of a JPEG encoder that exploits minimal usage of FPGA resources. The encoder compresses an image as a stream of 8×8 blocks with each element of the block applied and processed individually. The zigzag unit typically found in implementations of JPEG encoders is eliminated. The division operation of the quantization step is replaced by a combination of multiplication and shift operations. The encoder is implemented on Xilinx Spartan-3 FPGA and is benchmarked against two software implementations on four test images. It is demonstrated that it yields performance of similar quality while requiring very limited FPGA resources. A co-emulation technique is applied to reduce development time and to test and verify the encoder design.