Intraclass and between-class training-image registration for correlation-filter synthesis

Correlation filters have traditionally been designed without much attention given to the issue of the training images within a class or the relative spatial position between classes. We examine the impact of training-set registration on correlation-filter performance and develop techniques for centering the training images from a class that result in improved performance. We also show that it is beneficial to adjust the spatial position of the classes relative to one another. Although the proposed techniques are relevant for many types of correlation filter, we limit our discussion to algorithms for the maximum average correlation height filter and the distance classifier correlation filter.     

Two class minimax distance transform correlation filter

A new correlation filter formulation (that we refer to as the minimax distance transform correlation filter (MDTCF) is presented that minimizes the average squared distance from the filtered desired (or true-) class training images to a filtered reference image while maximizing the mean squared distance of the filtered undesired (or false-) class training images to this filtered reference image. This approach increases the separation between the false-class correlation outputs and the true-class correlation outputs. Classification can be performned using the squared distance of a filtered test image to the chosen filtered reference image. We show that the previously introduced distance classifier correlation filter (DCCF) is similar to a special case of MDTCF. We also examine the differences between the DCCF and the MDTCF, and show that MDTCF can offer increased discrimination performance. Also, MDTCF performance is evaluated on two different face databases.     

Distance-classifier correlation filters for multiclass target recognition

We describe a correlation-based distance-classifier scheme for the recognition and the classification of multiple classes. The underlying theory uses shift-invariant filters to compute distances between the input image and ideal references under an optimum transformation. The original distance-classifier correlation filter was developed for a two-class problem. We introduce a distance-classifier correlation filter that simultaneously considers multiple classes, and we show that the earlier two-class formulation is a special case of the classifier presented. Initial results are presented to demonstrate the discrimination- and distortion-tolerance capabilities of the proposed filter.     

Hybrid order statistic filter and its application to image restoration

We introduce a new nonlinear filter for signal and image restoration, the hybrid order statistic (HOS) filter. Because it exploits both rank- and spatial-order information, the HOS realizes the advantages of nonlinear filters in edge preservation and reduction of impulsive noise components while retaining the ability of the linear filter to suppress Gaussian noise. We show that the HOS filter exhibits improved performance over both the linear Wiener and the nonlinear L filters in reducing mean-squared error in the presence of contaminated Gaussian noise. In many cases it also performs favorably compared with the Ll and rank-conditioned rank selection filters.     

闪光CCD图像的中值-非线性扩散滤波

根据闪光CCD图像的特点,提出了一种中值-非线性扩散滤波(Median-NonlinearDiffusionFiltering,简称MNDF)方法。该方法采用中值预滤波来估计图像的真实边缘,通过求解偏微分方程(PartialDifferentialEquation,简称PDE)来进行非线性扩散滤波,充分发挥了中值滤波和非线性扩散滤波的优势,能更好地消除噪声、保护边缘。实验结果表明,在高斯噪声和脉冲噪声同时存在的情况下,MNDF方法取得的滤波效果较P-M方案和Catte方案要好,信噪比改善因子提高3~5倍,均方误差减小1.3~2.7倍。对闪光照相CCD图像取得了很好的消噪声结果,保护了边缘信息。     

Distortion-invariant pattern recognition with Fourier-plane nonlinear filters

The use of nonlinear techniques in the Fourier plane of pattern-recognition correlators can improve the correlators' performance in terms of discrimination against objects similar to the target object, correlation-peak sharpness, and correlation noise robustness. Additionally, filter designs have been proposed that provide the linear correlator with invariance properties with respect to input-signal distortions and rotations. We propose simple modifications to presently known distortion-invariant correlator filters that enable these filter designs to be used in a nonlinear correlator architecture. These Fourier-plane nonlinear filters can be implemented electronically, or they may be implemented optically with a nonlinear joint transform correlator. Extensive simulation results are presented that illustrate the performance enhancements that are gained by the unification of nonlinear techniques with these filter designs.     

一种新的自适应非线性卡尔曼滤波算法

为避免由于系统噪声统计特性不准确所导致的滤波性能下降问题,改进了一种基于新息的系统噪声方差调整方法,并将其与扩展卡尔曼滤波、Unscented 卡尔曼滤波和差分滤波相结合,形成自适应非线性卡尔曼滤波.将此方法应用到非线性测量光电跟踪系统中,并与采用基本非线性卡尔曼滤波进行性能对比.仿真实验结果证明该方法可以实时调整系统噪声方差,有效地避免由于系统噪声统计特性不准确所带来的滤波性能下降的问题,而且其性能明显优于基本非线性卡尔曼滤波.     

Rotationally invariant pattern recognition by use of linear and nonlinear cascaded filters

We discuss the merits of using single-layer (linear and nonlinear) and multiple-layer (nonlinear) filters for rotationally invariant and noise-tolerant pattern recognition. The capability of each approach is considered with reference to a two-class, rotation-invariant, character recognition problem. The minimum average correlation energy (MACE) filter is a linear filter that is generally accepted to be optimal for detecting signals that are free from noise. Here it is found that an optimized MACE filter cannot differentiate between the characters E and F in a rotation-invariant manner. We have found, however, that this task is possible when a single optimized linear filter is used to achieve the required response when a nonlinear threshold function is included after the filter. We show that this structure can be cascaded to form a multiple-layer, cascaded filter and that the capability of such a system is enhanced by its increased noise tolerance in the character recognition problem. Finally, we show the capability of a two-layer cascade as a means to detect different species of bacteria in images obtained from a phase-contrast microscope.     

Optical correlation of images with signal-dependent noise using constrained-modulation filter devices

Images with signal-dependent noise present challenges beyond those of images with additive white or colored signal-independent noise in terms of designing the optimal 4-? correlation filter that maximizes correlation-peak signal-to-noise ratio, or combinations of correlation-peak metrics. Determining the proper design becomes more difficult when the filter is to be implemented on a constrained-modulation spatial light modulator device. The design issues involved for updatable optical filters for images with signal-dependent film-grain noise and speckle noise are examined. It is shown that although design of the optimal linear filter in the Fourier domain is impossible for images with signal-dependent noise, proper nonlinear preprocessing of the images allows the application of previously developed design rules for optimal filters to be implemented on constrained-modulation devices. Thus the nonlinear preprocessing becomes necessary for correlation in optical systems with current spatial light modulator technlogy. These results are illustrated with computer simulations of images with signal-dependent noise correlated with binary-phase-only filters and ternary-phase-amplitude filters.     

Cascaded linear shift-invariant processors in optical pattern recognition

We study a cascade of linear shift-invariant processing modules (correlators), each augmented with a nonlinear threshold as a means to increase the performance of high-speed optical pattern recognition. This configuration is a special class of multilayer, feed-forward neural networks and has been proposed in the literature as a relatively fast best-guess classifier. However, it seems that, although cascaded correlation has been proposed in a number of specific pattern recognition problems, the importance of the configuration has been largely overlooked. We prove that the cascaded architecture is the exact structure that must be adopted if a multilayer feed-forward neural network is trained to produce a shift-invariant output. In contrast with more generalized multilayer networks, the approach is easily implemented in practice with optical techniques and is therefore ideally suited to the high-speed analysis of large images. We have trained a digital model of the system using a modified backpropagation algorithm with optimization using simulated annealing techniques. The resulting cascade has been applied to a defect recognition problem in the canning industry as a benchmark for comparison against a standard linear correlation filter, the minimum average correlation energy (MACE) filter. We show that the nonlinear performance of the cascade is a significant improvement over that of the linear MACE filter in this case.     

Distortion-tolerant three-dimensional object recognition with digital holography

We present a technique to implement three-dimensional (3-D) object recognition based on phase-shift digital holography. We use a nonlinear composite correlation filter to achieve distortion tolerance. We take advantage of the properties of holograms to make the composite filter by using one single hologram. Experiments are presented to illustrate the recognition of a 3-D object in the presence of out-of-plane rotation and longitudinal shift along the z axis.     

Spectral optimized asymmetric segmented phase-only correlation filter

We suggest a new type of optimized composite filter, i.e., the asymmetric segmented phase-only filter (ASPOF), for improving the effectiveness of a VanderLugt correlator (VLC) when used for face identification. Basically, it consists in merging several reference images after application of a specific spectral optimization method. After segmentation of the spectral filter plane to several areas, each area is assigned to a single winner reference according to a new optimized criterion. The point of the paper is to show that this method offers a significant performance improvement on standard composite filters for face identification. We first briefly revisit composite filters [adapted, phase-only, inverse, compromise optimal, segmented, minimum average correlation energy, optimal trade-off maximum average correlation, and amplitude-modulated phase-only (AMPOF)], which are tools of choice for face recognition based on correlation techniques, and compare their performances with those of the ASPOF. We illustrate some of the drawbacks of current filters for several binary and grayscale image identifications. Next, we describe the optimization steps and introduce the ASPOF that can overcome these technical issues to improve the quality and the reliability of the correlation-based decision. We derive performance measures, i.e., PCE values and receiver operating characteristic curves, to confirm consistency of the results. We numerically find that this filter increases the recognition rate and decreases the false alarm rate. The results show that the discrimination of the ASPOF is comparable to that of the AMPOF, but the ASPOF is more robust than the trade-off maximum average correlation height against rotation and various types of noise sources. Our method has several features that make it amenable to experimental implementation using a VLC.     

Optical threshold-decomposed Boolean-based gray-scale filter

Because of the advances made in optical technology, an extension can be made from the well-known stack filter to a new kind of gray-scale filter by using arbitrary Boolean functions instead of positive Boolean functions in the second step of a three-step process. Applications of these threshold-decomposed Boolean-based gray-scale filters in image processing include image smoothing and feature detection. A general optical implementation approach of nonlinear threshold-decomposed Boolean-based gray-scale filters is also described that utilizes area coding, gray scale correlation and lookup table thresholding techniques. Some experimental results are demonstrated.     

Filtering of Randomly Sampled Time-Stamped Measurements

This paper concerns the filtering of measurements that are taken by networked sensors at nonuniform intervals but that are accurately time stamped. Traditional digital filtering methods are difficult or impossible to use due to nonuniform sampling. Two filtering methods are described. Both are based on making an assumption about the signal behavior between measurements, such as the signal being constant between measurements. In the first method, a filter is formulated as an ordinary differential equation that is incrementally solved as measurements arrive. Such filtering is general; nonlinear and nontime invariant filters may be constructed. In the second method, signal convolution with a continuous-time finite impulse response filter is efficiently performed using a spline representation for the filter response. Such filters are ldquoFIR likerdquo in the sense that they have frequency-domain performance similar to FIR filters and have only slightly worse asymptotic computation time and memory requirements compared to FIR filters, yet have the advantage of being able to deal with nonuniformly sampled measurements. Examples of the operation of both sorts of filters are shown on actual measured data.     

Geometric approach for designing optical binary amplitude and binary phase-only filters

An analytic solution for real optimal filters is known, and the special case of optimal binary phase-only filters can be solved by a fast binning algorithm but no analytic solution is known. We establish a geometric solution for the design of optimal binary amplitude filters (OBAF's) and optimal binary phase-only filters (OBPOF's) for any object. The optimal filter is found in terms of maximizing the field strength at the origin in the correlation plane. We found that it is possible to construct a unique convex polygon by using an ordered set of phasors from the filter object's Fourier transform. This process leads eventually to an exact solution for the filter-design problem. We show that the maximum distance across the polygon divides the phasors into two groups: For the OBAF, it determines the group that is passed or blocked; for the OBPOF, it determines which group is passed with a zero or a pi phase shift. The shape of the convex polygon gives qualitative information on the criticalness and the tightness needed in the design process. It provides good insight into the binning-process algorithm and permits us to bound the error in the binning process. Design examples through computer simulation and applications in fingerprint identification are presented.     

Sequential Monte Carlo filters for abruptly changing state estimation

Sequential Monte Carlo techniques are evaluated for the nonlinear Bayesian filtering problem applied to systems exhibiting rapid state transitions. When systems show a large disparity between states (long periods of random diffusion about states interspersed with relatively rapid transitions), sequential Monte Carlo methods suffer from the problem known as sample impoverishment. In this paper, we introduce the maximum entropy particle filter, a new technique for avoiding this problem. We demonstrate the effectiveness of the proposed technique by applying it to highly nonlinear dynamical systems in geosciences and econometrics and comparing its performance with that of standard particle-based filters such as the sequential importance resampling method and the ensemble Kalman filter.     

Wavelet packet correlation methods in biometrics

We introduce wavelet packet correlation filter classifiers. Correlation filters are traditionally designed in the image domain by minimization of some criterion function of the image training set. Instead, we perform classification in wavelet spaces that have training set representations that provide better solutions to the optimization problem in the filter design. We propose a pruning algorithm to find these wavelet spaces by using a correlation energy cost function, and we describe a match score fusion algorithm for applying the filters trained across the packet tree. The proposed classification algorithm is suitable for any object-recognition task. We present results by implementing a biometric recognition system that uses the NIST 24 fingerprint database, and show that applying correlation filters in the wavelet domain results in considerable improvement of the standard correlation filter algorithm.     

Noise robustness of nonlinear filters for image recognition

We analyze the performance of the Fourier plane nonlinear filters in terms of signal-to-noise ratio (SNR). We obtain a range of nonlinearities for which SNR is robust to the variations in input-noise bandwidth. This is shown both by analytical estimates of the SNR for nonlinear filters and by experimental simulations. Specifically, we analyze the SNR when Fourier plane nonlinearity is applied to the input signal. Using the Karhunen-Loève series expansion of the noise process, we obtain precise analytic expressions of the SNR for Fourier plane nonlinear filters in the presence of various types of additive-noise processes. We find a range of nonlinearities that need to be applied that keep the output SNR of the filter stable relative to changes in the noise bandwidth.     

Spatial filtering velocimetry of objective speckles for measuring out-of-plane motion

This paper analyzes the dynamics of objective laser speckles as the distance between the object and the observation plane continuously changes. With the purpose of applying optical spatial filtering velocimetry to the speckle dynamics, in order to measure out-of-plane motion in real time, a rotational symmetric spatial filter is designed. The spatial filter converts the speckle dynamics into a photocurrent with a quasi-sinusoidal response to the out-of-plane motion. The spatial filter is here emulated with a CCD camera, and is tested on speckles arising from a real application. The analysis discusses the selectivity of the spatial filter, the nonlinear response between speckle motion and observation distance, and the influence of the distance-dependent speckle size. Experiments with the emulated filters illustrate performance and potential applications of the technology.     

Implementation of arbitrary real-valued correlation filters for the shadow-casting incoherent correlator

We describe an incoherent correlator, based on the shadow-casting principle, that is able to implement any real-valued linear correlation filter. The correlation filter and the input image are displayed on commercial liquid-crystal television (LCTV) panels. Although it cannot handle high-resolution images, the incoherent correlator is lensless, compact, low cost, and uses a white-light source. A bipolar technique is devised to represent any linear filter, computed from a single reference image or composite, in the correlator. We demonstrate experimentally the efficiency of the design in the case of optimal trade-off (OT) filters and optimal trade-off synthetic discriminant function (OT-SDF) filters.