Quasi-Reduced-State Soft-Output Viterbi Detector for Magnetic Recording Read Channel

Reduced-state trellis detection with decision feedback is widely used to reduce the energy consumption of trellis detectors, particularly for soft-output trellis detectors that are energy-hungry by nature. However, the decision feedback tends to increase the circuit critical path and, more important, makes it difficult to apply some well-proven high-speed trellis detector design techniques such as bit-level pipelining. This paper presents a method, referred to as quasi-reduced-state trellis detection, to tackle such speed bottlenecks. The basic idea is to simply obviate the use of decision feedback by mapping only the data storage block of the trellis detector onto a reduced-state trellis and keeping the trellis state metric computation on the original full-state trellis. This makes sense because the data storage block tends to dominate the overall energy consumption while the decision feedback is due to the reduced-state trellis metric computation. Therefore, it is intuitive that such quasi-reduced-state detectors may largely maintain the energy saving potentials of reduced-state trellis detection without being subject to decision-feedback-induced speed bottlenecks. We demonstrated the effectiveness of this proposed design method by using soft-output Viterbi algorithm (SOVA) detection for a magnetic recording read channel as a test vehicle.     

Parallel image restoration with a two-dimensional likelihood-based algorithm

We describe a pixelwise parallel algorithm for the restoration of images that have been corrupted by a low-pass optical channel and additive noise. This new algorithm is based on an iterative soft-decision method of error correction (i.e., turbo decoding) and offers performance on binary-valued imagery that is comparable to the Viterbi algorithm. We quantify the restoration performance of this new algorithm on random binary imagery for which it is superior to both the Wiener filter and the projection onto convex sets algorithms over a wide range of channels. For typical optical channels, the new algorithm is within 0.5 dB of the two-dimensional Viterbi restoration method [J. Opt. Soc. Am. A 17, 265 (2000)]. We also demonstrate the extension of our new algorithm to correlated and gray-scale images using vector quantization to mitigate the associated complexity burden. A highly parallel focal-plane implementation is also discussed, and a design study is presented to quantify the capabilities of such a VLSI hardware solution. We find that video-rate restoration on 252 x 252 pixel images is possible using current technology.     

Area-adaptive simulated annealing for image reconstruction and restoration

Simulated annealing (SA) is a robust, stable, but computationally costly method for solving ill-posed image-restoration problems. We describe the use of a backprojection operator that identifies those regions of an object estimate that have the greatest likelihood of being in error at each step of the SA process. This reduces computational time by concentrating the computing effort of SA on those pixels most effective in reducing the reconstruction error. The performance of an area-adaptive SA algorithm is evaluated for the restoration of images blurred by a simple pillbox space-invariant and a biconical space-variant point-spread function typical of a depth-measuring optical system.     

Cone-beam reconstruction by backprojection and filtering

A new analytical method for tomographic image reconstruction from cone-beam projections acquired on the source orbits lying on a cylinder is presented. By application of a weighted cone-beam backprojection, the reconstruction problem is reduced to an image-restoration problem characterized by a shift-variant point-spread function that is given analytically. Assuming that the source is relatively far from the imaged object, a formula for an approximate shift-invariant inverse filter is derived; the filter is presented in the Fourier domain. Results of numerical experiments with circular and helical orbits are considered.     

Digital restoration of defocused images in the wavelet domain

Many cameras nowadays are equipped with an autofocus mechanism that attempts to take pictures at the best possible focus. However, there are situations in which the pictures are still out of focus, such as when the photographer has mistakenly focused at a wrong position or when the focusing region consists of objects of different depths and therefore confuses the autofocus system. With a digital camera, we can attempt to use digital image-restoration techniques to bring the pictures back into focus. We design an algorithm that restores the image by digitally enhancing the corresponding frequency bands. We employ the restoration in the wavelet domain so that this restoration scheme can be compliant with JPEG 2000, which is positioned to succeed JPEG as the next image-compression standard and has the potential to be widely adopted by the digital photography industry owing to its many advanced features.     

Super-orthogonal space-time-frequency trellis coded OFDM

Super-orthogonal space-time trellis codes (SOSTTCs) given in the literature are full rate and provide full spatial diversity with high coding gain in narrowband quasistatic fading channels. The high number of parallel transitions in their trellis structure restricts their performance in wideband channels where the code performance suffers from multipath. Code design criteria are derived and a computer search based method is proposed to design full-rate optimised SOSTTCs exploiting full spatial and multipath diversity for multiple-input multiple-output orthogonal frequency division multiplexing systems. The proposed codes have codewords defined in space, time, and frequency. We evaluate the codeword error rates of the new 16, 32 and 64-state super-orthogonal space-time-frequency trellis codes for quadrature phase shift keying by computer simulation and show that they provide significant error performance improvement compared to their counterparts with equivalent delay length     

Upscaling image resolution of compact imaging systems using wavefront coding and a property of the point-spread function

We propose an image-resolution upscaling method for compact imaging systems. The image resolution is calculated using the resolving power of the optics and the pixel size of a digital image sensor. The resolution limit of the compact imaging system comes from its size and the number of allowed lenses. To upscale the image resolution but maintain the small size, we apply wavefront coding and image restoration. Conventional image restoration could not enhance the image resolution of the sensor. Here, we use the upscaled image of a wavefront-coded optical system and apply an image-restoration algorithm using a more precisely calculated point-spread function (PSF) as the deconvolution filter. An example of a wavefront-coded optical system with a 5-megapixel image sensor is given. The final image had a resolution equivalent to that of a 10-megapixel image using only four plastic lenses. Moreover, image degradation caused by hand motion could also be reduced using the proposed method.     

Rate 8/9 sliding block distance-enhancing code with stationarydetector

A new distance-enhancing code for partial-response magnetic recording channels eliminates most frequent errors, while keeping the two-step code trellis time invariant. Recently, published trellis codes either have lower code rates or result in time-varying trellises with a period of nine, thus requiring a higher complexity of detectors and code synchronization. The new code introduces dependency between code words in order to achieve the same coding constraints as the 8/9 time-varying maximum transition runlength (TMTR) code, with the same code rate, but resulting in a trellis that has a period of 2. This code has been applied to the E²PR4 and a 32-state generalized partial response (GPR) ISI target. The resulting two-step trellises have 14 and 28 states, respectively. Coding gain is demonstrated for both targets in additive white Gaussian noise     

高分辨率合成孔径雷达图像舰船检测方法

寻找针对高分辨率SAR图像的舰船目标检测算法。利用KSW双阈值分割技术，其效果比传统检测方法好，有利于进一步的目标分类和识别。且必须根据SAR图像分辨率来选择舰船检测算法。     

Scene-based nonuniformity correction with video sequences and registration

We describe a new, to our knowledge, scene-based nonuniformity correction algorithm for array detectors. The algorithm relies on the ability to register a sequence of observed frames in the presence of the fixed-pattern noise caused by pixel-to-pixel nonuniformity. In low-to-moderate levels of nonuniformity, sufficiently accurate registration may be possible with standard scene-based registration techniques. If the registration is accurate, and motion exists between the frames, then groups of independent detectors can be identified that observe the same irradiance (or true scene value). These detector outputs are averaged to generate estimates of the true scene values. With these scene estimates, and the corresponding observed values through a given detector, a curve-fitting procedure is used to estimate the individual detector response parameters. These can then be used to correct for detector nonuniformity. The strength of the algorithm lies in its simplicity and low computational complexity. Experimental results, to illustrate the performance of the algorithm, include the use of visible-range imagery with simulated nonuniformity and infrared imagery with real nonuniformity.     

Adaptive affinity propagation with spectral angle mapper for semi-supervised hyperspectral band selection

Band selection is a commonly used approach for dimensionality reduction in hyperspectral imagery. Affinity propagation (AP), a new clustering algorithm, is addressed in many fields, and it can be used for hyperspectral band selection. However, this algorithm cannot get a fixed number of exemplars during the message-passing procedure, which limits its uses to a great extent. This paper proposes an adaptive AP (AAP) algorithm for semi-supervised hyperspectral band selection and investigates the effectiveness of distance metrics for improving band selection. Specifically, the exemplar number determination algorithm and bisection method are addressed to improve AP procedure, and the relations between selected exemplar numbers and preferences are established. Experiments are conducted to evaluate the proposed AAP-based band selection algorithm, and the results demonstrate that the proposed method outperforms other popular methods, with lower computational cost and robust results.     

Nonseparable two-dimensional fractional fourier transform

Previous generalizations of the fractional Fourier transform to two dimensions assumed separable kernels. We present a nonseparable definition for the two-dimensional fractional Fourier transform that includes the separable definition as a special case. Its digital and optical implementations are presented. The usefulness of the nonseparable transform is justified with an image-restoration example.     

Noise-cancellation-based nonuniformity correction algorithm for infrared focal-plane arrays

The spatial fixed-pattern noise (FPN) inherently generated in infrared (IR) imaging systems compromises severely the quality of the acquired imagery, even making such images inappropriate for some applications. The FPN refers to the inability of the photodetectors in the focal-plane array to render a uniform output image when a uniform-intensity scene is being imaged. We present a noise-cancellation-based algorithm that compensates for the additive component of the FPN. The proposed method relies on the assumption that a source of noise correlated to the additive FPN is available to the IR camera. An important feature of the algorithm is that all the calculations are reduced to a simple equation, which allows for the bias compensation of the raw imagery. The algorithm performance is tested using real IR image sequences and is compared to some classical methodologies.     

Alternative implementations of the Two-Mu algorithm

I describe implementations of the Two-Mu image-restoration algorithm that model the center portion of the convolution of the point-spread function and the original image (this has been done heretofore), as well as those that model the full range of that convolution. The full convolution methods produce processed images of simple, simulated scenes that are comparable in quality with, and often involve computations that are considerably shorter than, those of the center convolution methods. The full convolution methods incur some loss of information near the edge of the scene. However, that loss may not be significant for large images, especially for those in which the important information is far from the edge of the scene.     

Detection of gaseous plumes in IR hyperspectral images using hierarchical clustering

The emergence of IR hyperspectral sensors in recent years enables their use in remote environmental monitoring of gaseous plumes. IR hyperspectral imaging combines the unique advantages of traditional remote sensing methods such as multispectral imagery and nonimaging Fourier transform infrared spectroscopy, while eliminating their drawbacks. The most significant improvement introduced by hyperspectral technology is the capability of standoff detection and discrimination of effluent gaseous plumes without need for a clear reference background or any other temporal information. We introduce a novel approach for detection and discrimination of gaseous plumes in IR hyperspectral imagery using a divisive hierarchical clustering algorithm. The utility of the suggested detection algorithm is demonstrated on IR hyperspectral images of the release of two atmospheric tracers. The application of the proposed detection method on the experimental data has yielded a correct identification of all the releases without any false alarms. These encouraging results show that the presented approach can be used as a basis for a complete identification algorithm for gaseous pollutants in IR hyperspectral imagery without the need for a clear background.     

Digital processing of electronic speckle pattern interferometry addition fringes

A digital image-processing method for analyzing double-pulsed electronic speckle pattern interferometry addition fringes is described. The procedure consists of three steps, forming a combination particularly suited to addressing some important practical limitations of the measurement system. In the first step it is shown that in certain cases fringe visibility may be enhanced by subtraction of a reference interferogram, so that a pattern with a quality similar to that of a subtraction one is obtained. In the second step noise is reduced by the application of a spectral subtraction image-restoration method. The third step concerns the calculation of the wrapped phase by means of a Fourier transform method with bandpass filtering. Preliminary experimental results that illustrate the performance of this approach are presented.     

Automated vehicle detection in forward-looking infrared imagery

We describe an algorithm for the detection and clutter rejection of military vehicles in forward-looking infrared (FLIR) imagery. The detection algorithm is designed to be a prescreener that selects regions for further analysis and uses a spatial anomaly approach that looks for target-sized regions of the image that differ in texture, brightness, edge strength, or other spatial characteristics. The features are linearly combined to form a confidence image that is thresholded to find likely target locations. The clutter rejection portion uses target-specific information extracted from training samples to reduce the false alarms of the detector. The outputs of the clutter rejecter and detector are combined by a higher-level evidence integrator to improve performance over simple concatenation of the detector and clutter rejecter. The algorithm has been applied to a large number of FLIR imagery sets, and some of these results are presented here.     

Kernel wavelet-Reed-Xiaoli: an anomaly detection for forward-looking infrared imagery

This paper describes a new kernel wavelet-based anomaly detection technique for long-wave (LW) forward-looking infrared imagery. The proposed approach called kernel wavelet-Reed-Xiaoli (wavelet-RX) algorithm is essentially an extension of the wavelet-RX algorithm (combination of wavelet transform and RX anomaly detector) to a high-dimensional feature space (possibly infinite) via a certain nonlinear mapping function of the input data. The wavelet-RX algorithm in this high-dimensional feature space can easily be implemented in terms of kernels that implicitly compute dot products in the feature space (kernelizing the wavelet-RX algorithm). In the proposed kernel wavelet-RX algorithm, a two-dimensional wavelet transform is first applied to decompose the input image into uniform subbands. A number of significant subbands (high-energy subbands) are concatenated together to form a subband-image cube. The kernel RX algorithm is then applied to this subband-image cube. Experimental results are presented for the proposed kernel wavelet-RX, wavelet-RX, and the classical constant false alarm rate (CFAR) algorithm for detecting anomalies (targets) in a large database of LW imagery. The receiver operating characteristic plots show that the proposed kernel wavelet-RX algorithm outperforms the wavelet-RX as well as the classical CFAR detector.     

Correlation-sensitive adaptive sequence detection

In high density magnetic recording, noise samples corresponding to adjacent signal samples are heavily correlated as a result of front-end equalizers, media noise, and signal nonlinearities combined with nonlinear filters to cancel them. This correlation significantly deteriorates the performance of detectors at high densities. In this paper, we propose a novel sequence detector that is correlation sensitive and adaptive to the nonstationary signal sample statistics. We derive the correlation-sensitive maximum likelihood detector. It can be used with any Viterbi-like receiver (e.g., partial response maximum likelihood, fixed delay tree search, multilevel decision feedback equalization) that relies on a tree/trellis structure. Our detector adjusts the metric computation to the noise correlation statistics. Because these statistics are nonstationary, we develop an adaptive algorithm that tracks the data correlation matrices. Simulation results are presented that show the applicability of the new correlation-sensitive adaptive sequence detector     

Multilevel spatial modulation

In this paper, a new Multilevel Spatial Modulation technique is proposed. It combines computationally efficient multilevel oding and spatial modulation based on trellis codes to increase coding gain, diversity gain, and bandwidth efficiency. The trellis complexity of the single-stage system increases exponentially, whereas in the proposed multilevel system the complexity increases linearly. The proposed system is analyzed with optimal Viterbi and suboptimal sequential decoding algorithms. The results show that sequential decoding saves 75% of the computational power with a loss of 2 dB SNR approximately, when compared with optimal Viterbi decoding, over both fast- and slow-fading channel conditions. Since the antenna index is used as a source of information in spatial modulation, the number of antennae required increases with the throughput and packing a large number of antennas make cross-correlation unavoidable. In this paper, a low complexity modified decoding technique is also proposed for the multilevel spatial modulation system, in which the correlated received signals are equally combined and decoded by the multistage decoder using the Viterbi algorithm. This technique exploits the receiver antenna correlation and makes the decoding complexity independent of number of antennas. The simulation results indicate that the proposed low complexity algorithm gives approximately 8–10 dB gain when compared with optimal Viterbi decoder with equivalent computational complexity when the eight highly correlated signals are equally combined. This may be a suitable solution for mobile handsets where size and computational complexity are the major issues.