A two-stage lossless image compression scheme is presented, the first stage of which decorrelates the raw image data by using a two-dimensional, causal linear predictor, resulting in significant entropy reduction. The second stage uses standard coding techniques. Different two-stage schemes are compared with respect to their performances on several images, using bilevel coding, arithmetic coding, and adaptive Huffman coding.

The problem of designing optimum Laguerre networks for nonrecursive discrete-time systems whose impulse responses are specified is studied. Optimization is achieved by minimizing the mean-square error between the impulse response of the given discrete system and its Laguerre approximation. An efficient algorithm for evaluating the coefficients of the polynomial whose roots yield the optimum solution is also presented.>

Describes a family of adaptive change detectors for detecting and segmenting moving objects in video scenes. These change detectors employ several techniques to allow them to adjust to changing image conditions. One of these is a dynamic reference frame which allows the system to adapt to changing backgrounds and produces an unambiguous target silhouette for classification. Another is a scheme to estimate system noise and adjust detection thresholds accordingly to satisfy a chosen performance criterion. Finally, adaptive normalization is used to maintain constant image contrast.>

This letter compares two parameter estimation schemes using the variance of the estimate as the criterion for comparison.

This paper presents an adaptive scheme for the detection and identification of the sinusoidal componments in a received signal. It uses a recursive ALE (Adaptive Line Enhancer) which is implemented as the sum of constrained second-order bandpass filters. Although not as robust as other non-recursive schemes for detection and identification, it is far more efficient with respect to the number of computations required.

The purpose of this paper is to present a simple adaptive algorithm for detecting and tracking a sinusoid in broadband noise, while at the same time improving its signal-to-noise ratio (SNR).

In this paper we present an adaptive short-term correlator (ASC) algorithm, and present two applications of the same. These applications are, (i) line enhancement of a single or dominant sinusoid, and, (ii) detection of sources that produce narrowband correlated signals at two sensors, over a period of time.

The main objective of this paper is to compare two adaptive algorithms for estimating the time delay between two-sensor outputs. These methods are very similar in the sense that they minimize the mean-squared error using a least mean square (LMS) gradient approach to estimate the time difference of arrival to two sensors. However, they apply the LMS gradient method in different ways, thereby holding different properties. Some theoretical aspects are addressed. Also, they are compared via computer simulations for a variety of cases, which include time-varying delay functions corresponding to moving source or receivers.

This paper concerns a study of some aspects of the SCOT (smoothed coherence transform) and Roth algorithms, which are useful for time delay estimation. The behaviour of the SCOT algorithm for the time delay estimation of sinusoids is investigated, and adaptive time domain implementations for the Roth and SCOT algorithms are presented.

This paper concerns an adaptive approach for estimateing the magnitude-squared coherence (MSC) function via Widrow's least-mean-square (LMS) algorithm. Some theoretical aspects are addressed, and simulation results are presented to evaluate the performance of the adaptive approach.

The purpose of this paper is two-fold. First, it is to present further simulation results concerning the generalized cross correlation (GCC) method fer estimating the time delay between received signals at two or more sensors (receivers) when the source is fixed. The second purpose is to discuss some preliminary results related to extending the GCC approach for time delay estimation for the case of a moving source. The specific cases of the GCC approach considered in this paper are the standard cross correlation (SCC) method, the smoothed coherence transform (SCOT) method, and the approximate maximum likelihood (AML) method.

A method for generating a delayed version of a given bandlimited signal is presented, where the delay could be time-varying. Simulation results which illustrate the performance of this method are also included.

A procedure is described for the computation of the discrete cosine transform (DCT) via the use of the arcsine transform. The approach eliminates time-consuming multiplications, the DCT being accomplished with only additions and table lookups. While a fast Fourier transform (FFT) approach to computing the DCT involves on the order of N\log_{2}2N butterfly computations to evaluate all N coefficients, the arcsine method requires only 4N - 1 real additions and 2N table lookups to evaluate each DCT coefficient. Thus, for applications in which M coefficients are desired or when N is reasonably small (say, N \leq 256 ), the arcsine approach is favored over that of the FFT. Some approaches to hardware implementation are presented.

In the interest of localizing underwater acoustic sources, several digital signal processing techniques have been proposed for the important intermediate step of time delay estimation. One such method is the SCOT (Smoothed Coherence Transform) method which is designed to counteract the undesirable effects of strong sinusoids. The purpose of this paper is to compare the performance of the SCOT with the standard cross correlation method when sinusoids are not present. Under these conditions it is shown that the cross correlation method outperforms the SCOT. The results so obtained are compared with some of those published recently.

The use of the sequential regression (SER) algorithm [1, 2] for long-term processing applications is limited by two problems which can occur when an SER predictor has more weights than required to predict the input signal. First, computational difficulties related to updating the autocorrelation matrix inverse could arise, since no unique least-squares solution exists. Second, the predictor strives to remove very low-level components in the input, and hence could implement a gain function that is essentially zero over the entire pass-band. The predictor would then tend to become a "no-pass" filter which is undesirable in certain applications -- e.g., intrusion detection [6]. Modifications to the SER algorithm that overcome the above problems are presented, which enable its use for long-term signal processing applications.

This paper presents a recurrence relation that enables one to obtain the inverse filter associated with an adaptive escalator predictor (AEP). The behaviour of this inverse filter is illustrated by means of a simple experiment related to the detection and frequency estimation of a sinusoidal burst embedded in band-limited noise.

In the paper, we introduce the notion of short-term adaptive filtering via a sequential regression (SER) formulation. A corresponding short-term SER algorithm for nonrecursive filters is derived. Experimental results involving a short-term SER predictor are presented. For purposes of comparison, corresponding results using Widrow's least-mean-square (LMS), and the conventional SER algorithm are also included.

A tutorial-review paper on discrete orthogonal transforms and their applications in digital signal and image (both monochrome and color) processing is presented. Various transforms such as discrete Fourier, discrete cosine, Walsh-Hadamard, slant, Haar, discrete linear basis, Hadamard-Haar, rapid, lower triangular, generalized Haar, slant Haar and Karhunen-Loêve are defined and developed. Pertinent properties of these transforms such as power spectra, cyclic and dyadic convolution and correlation are outlined. Efficient algorithms for fast implementation of these transforms based on matrix partitioning or matrix factoring are presented. The application of these transforms in speech and image processing, spectral analysis, digital filtering (linear, nonlinear, optimal and suboptimal), nonlinear systems analysis, spectrography, digital holography, industrial testing, spectrometric imaging, feature selection, and patter recognition is presented. The utility and effectiveness of these transforms are evaluated in terms of some standard performance criteria such as computational complexity, variance distribution, mean-square error, correlated rms error, rate distortion, data compression, classification error, and digital hardware realization.

A discrete cosine transform (DCT) is defined and an algorithm to compute it using the fast Fourier transform is developed. It is shown that the discrete cosine transform can be used in the area of digital processing for the purposes of pattern recognition and Wiener filtering. Its performance is compared with that of a class of orthogonal transforms and is found to compare closely to that of the Karhunen-Loève transform, which is known to be optimal. The performances of the Karhunen-Loève and discrete cosine transforms are also found to compare closely with respect to the rate-distortion criterion.

Cyclic and dyadic shifts of a periodic sequence are illustrated on the H diagram. Fast Walsh-Hadamard transform of the shifted (both cyclic and dyadic) sequence is represented graphically on the H diagram. The shift-invariant properties of the power spectrum are derived from these operations.

Using elementary matrix partitioning techniques, a fast algorithm for computing the complex BIFORE transform is developed.