Search

Multicolor fluorescence in situ hybridization (M-FISH) techniques provide color karyotyping that allows simultaneous analysis of numerical and structural abnormalities of whole human chromosomes. Chromosomes are stained combinatoriaily in M-FISH. By analyzing the intensity combinations of each pixel, all chromosome pixels in an image are classified. Due to the overlap of excitation and emission spectra and the broad sensitivity of image sensors, the obtained images contain crosstalk between the color channels. The crosstalk complicates both visual and automatic image analysis and may eventually affect the classification accuracy in M-FISH. The removal of crosstalk is possible by finding the color compensation matrix, which quantifies the color spillover between channels. However, there exists no simple method of finding the color compensation matrix from multichannel fluorescence images whose specimens are combinatorially hybridized. In this paper, we present a method of calculating the color compensation matrix for multichannel fluorescence images whose specimens are combinatorially stained. Index Terms--Chromosomes, color compensation, crosstalk, fluorescence in-situ hybridization (FISH), fluorescence, multicolor fluorescence in-situ hybridization (M-FISH), multicolor, multi-spectral.

Probability density functions for the components of the first and second derivatives of localized phase are derived.

The derivation of bounds on the structural similarity (SSIM) index as a function of quantization rate for fixed-rate uniform quantization is discussed. Findings reveal the importance of the proposed bounds for rate allocation problems in practical image and video coding applications.

Multicolor fluorescence in situ hybridization (M-FISH) techniques provide color karyotyping that allows simultaneous analysis of numerical and structural abnormalities of whole human chromosomes. Chromosomes are stained combinatorially in M-FISH. By analyzing the intensity combinations of each pixel, all chromosome pixels in an image are classified. Often, the intensity distributions between different images are found to be considerably different and the difference becomes the source of misclassifications of the pixels. Improved pixel classification accuracy is the most important task to ensure the success of the M-FISH technique. In this paper, we introduce a new feature normalization method for M-FISH images that reduces the difference in the feature distributions among different images using the expectation maximization (EM) algorithm. We also introduce a new unsupervised, nonparametric classification method for M-FISH images. The performance of the classifier is as accurate as the maximum-likelihood classifier, whose accuracy also significantly improved after the EM normalization. We would expect that any classifier will likely produce an improved classification accuracy following the EM normalization. Since the developed classification method does not require training data, it is highly convenient when ground truth does not exist. A significant improvement was achieved on the pixel classification accuracy after the new feature normalization. Indeed, the overall pixel classification accuracy improved by 20% after EM normalization. Index Terms--Chromosome, classification, expectation maximization (EM), maximum-likelihood, multicolor fluorescence in situ hybridization (M-FISH), normalization, unsupervised.

An algorithm for designing linear equalizers, which maximizes the structural similarity (SSIM) index between the reference and restored signals, is proposed. Findings reveal the achievement of higher SSIM index and superior perceptual quality.

A gaze-attentive fixation finding engine (GAFFE) that utilizes a bottom-up modality for fixation selection in natural scenes is presented to assess the contributions of four low-level local image features. The image regions, which are likely to need fixation, selected by the suggested method seem to match with the fixations noted from human observers.

A multilinear expansion of the probability density function of the source statistics is used to refine the classical independent component analysis (ICA) decomposition. The new enhanced multilinear ICA (MICA) model can be used to analyze nonstationarity natural images using natural scene statistics models.

A foveated, multifixating strategy is presented for locating corners in natural images. The results have shown that the algorithm performs well on strong edges with sharp corners and less well in areas of fine detail.

A performance evaluation study of ten image quality assessment (QA) algorithms are presented, which involved 779 distorted images derived from 29 source images using five different distortion types. The study was diverse in terms of image content, distortion types, distortion strength and the number of human subjects ranking each image, and results and data from the study are made publicly available to facilitate future research in image quality assessment.

The concept of extracting features of the original high-quality images and embedding into the image data as invisible hidden messages and can provide an objective measure of the quality of the distorted images known as quality-aware image is described. A practical quality-aware image encoding, decoding and quality analysis system is formulated which uses a unique reduced-reference image quality assessment technique based on a statistical model of natural images and quantization watermarking-based data hiding method in wavelet transform domain.

A statistical model of key compression techniques for estimating the distortion in progressive JPEG compressed images due to quantization and channel hit errors in the joint manner is presented. The unequal power allocation scheme used to prove the viability of the model showed that peak-signal-to-noise ratio (PSNR) increased at low SNRs when compared to equal power allocation.

The goal of quality assessment (QA) research is to design algorithms that can automatically assess the quality of images or videos in a perceptually consistent manner. The problem of image QA is approached by proposing a novel information fidelity criterion that is based on natural scene statistics.

Traditional chromosome imaging has been limited to grayscale images, but recently a 5-fluorophore combinatorial labeling technique (M-FISH) was developed wherein each class of chromosomes binds with a different combination of fluorophores. This results in a multispectral image, where each class of chromosomes has distinct spectral components. In this paper, we develop new methods for automatic chromosome identification by exploiting the multispectral information in M-FISH chromosome images and by jointly performing chromosome segmentation and classification. We 1) develop a maximum-likelihood hypothesis test that uses multispectral information, together with conventional criteria, to select the best segmentation possibility; 2) use this likelihood function to combine chromosome segmentation and classification into a robust chromosome identification system; and 3) show that the proposed likelihood function can also he used as a reliable indicator of errors in segmentation, errors in classification, and chromosome anomalies, which can be indicators of radiation damage, cancer, and a wide variety of inherited diseases. We show that the proposed multispectral joint segmentation-classification method outperforms past grayscale segmentation methods when decomposing touching chromosomes. We also show that it outperforms past M-FISH classification techniques that do not use segmentation information. Index Terms--Chromosomes, image segmentation, karyotyping, object recognition, partial occlusion.

A model that accounts for the properties of the imaging system and of the tissues is developed to get a better understanding of confocal reflectance microscopy images of cervical tissues, and to supply a basis for parameter selection in a classification algorithm. The algorithm was applied to an extensive set of cervical images and it was found that it detects 90% of hand-segmented nuclei with an average of 6 false positive per frame.

Theorems are developed that place limits on the point-wise approximation of the responses of filters, both linear shift invariant (LSI) and linear shift variant (LSV) to input signals and images that are LSV. The application of the theory was found to defoveation of images, deblurring of shift-variant blurs, and shift-variant edge detection.

A study of the objective method for assessing perceptual image quality that attempted to quantify the visibility of errors between distorted and reference images according to known characteristics of the human visual system (HVS) is presented. The development of the structural similarity index (SSIM) for quality assessment based on the assumption that HVS is highly adapted to extract structural information is discussed.

A wavelet-based, scalable video coding system, foveation scalable video coding (FSVC, is described and discussed.

This paper explores the problem of communicating high-quality, foveated video streams in real time. Foveated video exploits the nonuniform resolution of the human visual system by preferentially allocating bits according to the proximity to assumed visual fixation points, thus delivering perceptually high quality at greatly reduced bandwidths. Foveated video streams possess specific data density properties that can be exploited to enhance the efficiency of subsequent video processing. Here, we exploit these properties to construct several efficient foveated video processing algorithms: foveation filtering (local bandwidth reduction), motion estimation, motion compensation, video rate control, and video postprocessing. Our approach leads to enhanced computational efficiency by interpreting nonuniform-density foveated images on the uniform domain and by using a foveation protocol between the encoder and the decoder. Index Terms--Foveated video, foveated video algorithms, foveated video processing.

Blocking artifacts continue to be among the most serious defects that occur in images and video streams compressed to low bit rates using block discrete cosine transform (DCT)-based compression standards (e.g, JPEG, MPEG, and H.263). It is of interest to be able to numerically assess the degree of blocking artifact in a visual signal, for example, in order to objectively determine the efficacy of a compression method, or to discover the quality of video content being delivered by a web server. In this paper, we propose new methods for efficiently assessing, and subsequently reducing, the severity of blocking artifacts in compressed image bitstreams. The method is blind, and operates only in the DCT domain. Hence, it can be applied to unknown visual signals, and it is efficient since the signal need not be compressed or decompressed. In the algorithm, blocking artifacts are modeled as 2-D step functions. A fast DCT-domain algorithm extracts all parameters needed to detect the presence of, and estimate the amplitude of blocking artifacts, by exploiting several properties of the human vision system. Using the estimate of blockiness, a novel DCT-domain method is then developed which adaptively reduces detected blocking artifacts. Our experimental results show that the proposed method of measuring blocking artifacts is effective and stable across a wide variety of images. Moreover, the proposed blocking-artifact reduction method exhibits satisfactory performance as compared to other post-processing techniques. The proposed technique has low computational cost hence can be used for real-time image/video quality monitoring and control, especially in applications where it is desired that the image/video data be processed directly in the DCT-domain. Index Terms--Blocking artifacts, DCT-domain, human visual perception, image quality control, video post-processing.

The techniques of digital signal processing (DSP) and digital image processing (DIP) have found a myriad of applications in diverse fields of scientific, commercial, and technical endeavor. DSP and DIP education needs to cater to a wide spectrum of people from different educational backgrounds. This paper describes tools and techniques that facilitate a gentle introduction to fascinating concepts in signal and image processing. Novel LabVIEW- and MATLAB-based demonstrations are presented, which, when supplemented with Web-based class lectures, help to illustrate the power and beauty of signal and image-processing algorithms. Equipped with informative visualizations and a user-friendly interface, these modules are currently being used effectively in a classroom environment for teaching DSP and DIP at the University of Texas at Austin (UT-Austin). Most demonstrations use audio and image signals to give students a flavor of real-world applications of signal and image processing. This paper is also intended to provide a library of more than 50 visualization modules that accentuate the intuitive aspects of DSP algorithms as a free didactic tool to the broad signal and image-processing community. Index Terms--Demonstration library, interactive education, multidisciplinary, signal and image-processing education, visualization.

Discrete cosine transform (DCT) based digital video coding standards such as MPEG and H.26x are becoming more widely adopted for multimedia applications. Since the standards differ in their format and syntax, video transcoding, where a compressed video bit-stream is converted from one format to another format, is of interest for purposes such as channel bandwidth adaptation and video composition. DCT-domain video transcoding is generally more efficient than spatial-domain transcoding. However, since the data is organized block by block in the DCT domain, inverse motion compensation becomes the bottleneck for DCT-domain methods. In this paper, we propose a novel local bandwidth constrained fast inverse motion compensation algorithm operating in the DCT domain. Relative to Chang's algorithm, the proposed algorithm achieves computational improvement of 25%-55% without visual degradation. A by-product of the proposed algorithm is a reduction of blocking artifacts in very low bit-rate compressed video sequences. The proposed algorithm can be combined with other fast methods presented in the literature for more computational savings. We also present a look-up-table (LUT) based implementation method by modeling the statistical distribution of the DCT coefficients in natural images and video sequences. By this method, we obtain a further another 31%-48% improvement in computation. The memory requirement of the LUT is about 800 kB, which is reasonable. Moreover, the LUT can be shared by multiple DCT-domain video processing applications running on the same computer or video server. Index Terms--DCT-domain, inverse motion compensation, MPEG video, video transcoding, video composition.

We propose a new anisotropic diffusion filter for denoising low-signal-to-noise molecular images. This filter, which incorporates a median filter into the diffusion steps, is called an anisotropic median-diffusion filter. This hybrid filter achieved much better noise suppression with minimum edge blurring compared with the original anisotropic diffusion filter when it was tested on an image created based on a molecular image model. The universal quality index, proposed in this paper to measure the effectiveness of denoising, suggests that the anisotropic median-diffusion filter can retain adherence to the original image intensities and contrasts better than other filters. In addition, the performance of the filter is less sensitive to the selection of the image gradient threshold during diffusion, thus making automatic image denoising easier than with the original anisotropic diffusion filter. The anisotropic' median-diffusion filter also achieved good denoising results on a piecewise-smooth natural image and real Raman molecular images. Index Terms--Anisotropic diffusion, image enhancement, molecular imaging, nonlinear filter.

It is shown that formulating the nullspace-based approach to eigenstructure-based direct multichannel blind image restoration methods implies a subspace-based approach that relies on matrix operations. The new approach has the same advantages as the nullspace-based one but requires less computational complexity.

An embedded foveation image coding (EFIC) algorithm is proposed. The algorithm orders the encoded bitstream to optimize foveated visual quality at arbitrary bit rates.

AM-FM representation was used for fingerprint classification to obtain significant gains in performance in relation to the National Institute of Standards system.

Multidimensional multicomponent and quasi-Eigenfunction approximation AM-FM models are discussed. Techniques on several images of general interest in practical applications have been demonstrated. Reconstructions that establish validity of characterizing the type of images as sums of locally narrowband modulated components have been made. Two discrete computational paradigms were used. Dominant component analysis does one thing and channelized components analysis does another.

A study was conducted to characterize a system for calculating three-dimensional range segmentation associated with an arbitrary visible scene. The system utilizes focus information and determines range segmentation in three steps. It carries out range classification by evaluating focus cues within Bayesian estimation. The system also segments scenes into three-dimensional regions in range segmentation by associating connected object surfaces with various depth ranges.

We show that imposing a certain number of vanishing moments on a scaling function (e.g., coiflets) leads to fairly small phase distortion on its associated filter bank in the neighborhood of DC. However, the phase distortion at the other frequencies can be much larger. We design a new class of real-valued, compactly supported, orthonormal, and nearly symmetric wavelets (we call them generalized coiflets) with a number of nonzero-centered vanishing moments equally distributed on scaling function and wavelet. Such a generalization of the original coiflets offers one more free parameter, the mean of the scaling function, in designing filter banks. Since this parameter uniquely characterizes the first several moments of the scaling function, it is related to the phase response of the lowpass filter at low frequencies. We search for the optimal parameter to minimize the maximum phase distortion of the filter bank over the lowpass half-band. Also, we are able to construct nearly odd-symmetric generalized coiflets, whose associated lowpass filters are surprisingly similar to those of some biorthogonal spline wavelets. These new wavelets can be useful in a broad range of signal and image processing applications because they provide a better tradeoff between the two desirable but conflicting properties of the compactly supported and real-valued wavelets, i.e., orthonormality versus symmetry, than the original coiflets. Index Terms - FIR digital filters, quadrature mirror filters, wavelet transforms.

A study was conducted on a new approach to image estimation based on a flexible constraint framework that shows meaningful structural image assumptions. The findings indicate that approximate solutions may be computed in reasonable time using the novel generalized deterministic annealing optimization technique. Moreover, the findings demonstrate high quality image estimation as measured by local feature integrity, and improvement in signal-to-noise ratio.

An analysis was presented to evaluate the accuracy of generalized coiflet-based sampling approximation of smooth functions by developing the convergence rates for local approximation error. The method used was computational analysis of data. Results revealed that generalized coiflets can achieve asymptotically better approximation than conventional coiflets and approximation of expansion coefficients by function samples is asymptotically negligible.

The asymptotic behavior of the dual filters related with biorthogonal spline wavelet (BSW) systems and general biorthogonal Coifman wavelet (GBCW) systems were analyzed. Results indicated that magnitude responses of the dual filters in BSW systems diverge or converge to frequency responses. On the other hand, synthesis filters in the GBCW systems converge to halfband lowpass filter.

Theorems for analyzing discrete AM-FM functions and their behavior in discrete linear systems and simple discrete nonlinear systems comprising products of discrete linear subsystems are examined. Quasieigenfunction approximations were derived that complement the eigenfunction interpretation of sinusoids in linear systems. The approximations are found to be effective in analyzing discrete-time nonlinear energy operators and can lead to the selection of an optimal class of bandpass filters for a discrete multiband AM-FM demodulation system.

A statistical component model enables computation of AM-FM representations for multicomponent, nonstationary images. A filterbank consisting of frequency and orientation selective channels isolates the components. Localized nonlinear operators and optimal MMSE estimators give estimates of the modulating functions for each component. The operators use the channel responses for the estimations.

We develop a general formalism for computing high quality, low-cost solutions to nonconvex combinatorial optimization problems expressible as distributed interacting local constraints. For problems of this type, generalized deterministic annealing (GDA) avoids the performance-related sacrifices of current techniques. GDA exploits the localized structure of such problems by assigning K-state neurons to each optimization variable. The neuron values correspond to the probability densities of K-state local Markov chains and may be updated serially or in parallel; the Markov model is derived from the Markov model of simulated annealing (SA), although it is greatly simplified. Theorems are presented that firmly establish the convergence properties of GDA, as well as supplying practical guidelines for selecting the initial and final temperatures in the annealing process. A benchmark image enhancement application is provided where the performance of GDA is compared to other optimization methods. The empirical data taken in conjunction with the formal analytical results suggest that GDA enjoys significant performance advantages relative to current methods for combinatorial optimization.

A multiband or wavelet approach helps obtain AM-FM information from noisy modulated signals. A nonlinear energy operator aids the isolation of AM-FM energy which is subjected to an energy separation algorithm. Signal clarity is further enhanced by filtering with multiple bandpass filters. The major local channel response helps the analysis of the emerging signal, at every instant. The use of constant-Q or multiscale wavelet-like filter bank ensures uniform performance across bands.

A measure for the smoothness of a signal is provided by the study of the group of root signals of the central filter which displays the concept of local monotonicity. A regionally monotonic output is not obtained from the filter pass and hence the central filter is a suboptimal smoother under a measure of smoothness. Even in the presence of a regionally monotonic output, it may or may not possess the required properties. Locally monotonic regression is an efficient method for the optimal smoothing of finite-length discrete signals.

Visual pattern image sequence coding (VPISC) is a new digital image sequence (video) coding process that possesses significant advantages relative to other technologies; in particular, it is extremely efficient in terms of the computational effort required. It is designed to exploit properties of the human visual systems (HVS), and thus yields high visual fidelity. Visual quality criteria are deliberately chosen over information-theoretic ones on the grounds that, in images intended for human viewing, visual criteria are the most meaningful ones. VPISC yields impressive compression comparable to other recent methods, such as motion-compensated vector quantization. VPISC divides the images into spatiotemporal cubes, which are then independently matched with one of a small, predetermined set of visually meaningful three-dimensional space-time patterns. The pattern set is chosen to conform to specific characteristics of the HVS. Also introduced are two modifications of VPISC: adaptive and foveal VPISC. These are spatiotemporally nonuniform implementations that code different portions of the image sequence at different resolutions, according to either a fidelity criterion (for AVPISC) or a foveation criterion (for FVPISC).

Abstract The increasing number of demanding consumer video applications, as exemplified by cell phone and other low-cost digital cameras, has boosted interest in no-reference objective image and video quality assessment (QA) algorithms. In this paper, we focus on no-reference image and video blur assessment. We consider natural scenes statistics models combined with multi-resolution decomposition methods to extract reliable features for QA. The algorithm is composed of three steps. First, a probabilistic support vector machine (SVM) is applied as a rough image quality evaluator. Then the detail image is used to refine the blur measurements. Finally, the blur information is pooled to predict the blur quality of images. The algorithm is tested on the LIVE Image Quality Database and the Real Blur Image Database; the results show that the algorithm has high correlation with human judgments when assessing blur distortion of images.

The asymmetrical filterbanks (AFBs) for image coding applications proposed by O. Egger and W. Li are not new since their proposed construction of AFBs is similar to the factorization of Lagrange halfband filters, which has been reported in the earlier work by R. Ansari et al. and by A. Cohen et al. Moreover, one of the AFBs has already been applied to wavelet transform-based image coding by Antonini et al. Egger and Li also made an error in formulating the construction of AFBs.

The regression method, widely used in statistics, is used to filter and shape signals. Isotonic regression and linear regression, which are used in statistics along with monotonic regression are applied in signal processing. The regression techniques are also useful for analyzing the signal smooth paradigm, which is subjected to deterministic and geometric concepts of projection.

Studies of nonlinear signal operators developed by Teager on speech modeling reveal conditions that ensure that the Teager-Kaiser energy operator for continuous-time signals and the output of the respective discrete-time energy operator is uniformly positive. The relative error is minor for realistic signals. The conditions necessary for positivity to be maintained are expressed in terms of logarithm signal magnitude concavity.

A proposed integral inequality limiting the weighted integrated absolute deviation on n-dimensional real function which is differentiable over an interval is tested. The integral inequality is assumed dependent on any value equated with the function within the interval. Results show that the inequality can be implemented in specific set-theoretic programs for restoring signals.

Natural scene statistics (NSS) is used for blindly measuring the quality of images compressed by JPEG2000 image coder. An average error of 8.05 is achieved from the mean opinion scores.

This paper describes the application of an amplitude modulation-frequency modulation (AM-FM) image representation in segmenting electron micrographs of skeletal muscle for the recognition of: 1) normal sarcomere ultrastructural pattern and 2) abnormal regions that occur in sarcomeres in various myopathies. A total of 26 electron micrographs from different myopathies were used for this study. It is shown that the AM-FM image representation can identify normal repetitive structures and sarcomeres, with a good degree of accuracy. This system can also detect abnormalities in sarcomeres which alter the normal regular pattern, as seen in muscle pathology, with a recognition accuracy of 75%-84% as compared to a human expert. Index Terms--AM-FM modeling, electron microscopy, muscle imaging, myopathies, texture analysis.

Foveation has been investigated as a tool for exploring techniques for optimal rate control of foveated video. It is concluded that foveated compression and rate control offers significant potential with regard to compression performance, speed and visual quality.

Partial differential equations and AM-FM image modeling was used to achieve an automated method for repairing broken, occluded oriented image textures.

A new class of multidimensional orthogonal FM transforms is presented. It has been possible to show promising coding gains, particularly on images with noiselike, rough or granular textured regions.

A fast, high quality inverse halftoning algorithm is presented for error diffused halftones.

Compaction by optimal permutation or COPERM is used for transform-domain energy compaction of broadband signals. The concept behind COPERM is that any signal can be transformed to become a more desirable signal from a class of target signals through a suitable permutation of its samples. Transform-domain energy compaction is applied to lossy compression, and some broadband image compression results are presented.