Poisson inverse problems by the Plug-and-Play scheme

The easy-to-compute Anscombe transform offers a conversion of a Poisson random variable into a variance stabilized Gaussian one, thus becoming handy in various Poisson-noisy inverse problems. Solution to such problems can be done by applying this transform, then invoking a high-performance Gaussian-noise-oriented restoration algorithm, and finally using an inverse transform. This process works well for high-SNR images, but when the noise level is high, it loses much of its effectiveness. This work suggests a novel method for coupling Gaussian denoising algorithms to Poisson noisy inverse problems. This approach is based on a general approach termed “Plug-and-Play-Prior”. Deploying this to Poisson inverse-problems leads to an iterative scheme that repeats an easy treatable convex programming task, followed by a powerful Gaussian denoising This method, like the Anscombe transform, enables to plug Gaussian denoising algorithms for the Poisson-oriented problem, and yet, it is effective for all SNR ranges.     

A statistical multiscale framework for Poisson inverse problems

This paper describes a statistical multiscale modeling and analysis framework for linear inverse problems involving Poisson data. The framework itself is founded upon a multiscale analysis associated with recursive partitioning of the underlying intensity, a corresponding multiscale factorization of the likelihood (induced by this analysis), and a choice of prior probability distribution made to match this factorization by modeling the “splits” in the underlying partition. The class of priors used here has the interesting feature that the “noninformative” member yields the traditional maximum-likelihood solution; other choices are made to reflect prior belief as to the smoothness of the unknown intensity. Adopting the expectation-maximization (EM) algorithm for use in computing the maximum a posteriori (MAP) estimate corresponding to our model, we find that our model permits remarkably simple, closed-form expressions for the EM update equations. The behavior of our EM algorithm is examined, and it is shown that convergence to the global MAP estimate can be guaranteed. Applications in emission computed tomography and astronomical energy spectral analysis demonstrate the potential of the new approach     

Remote sensing image registration via active contour model

Image registration is the process by which we determine a transformation that provides the most accurate match between two images. The search for the matching transformation can be automated with the use of a suitable metric, but it can be very time-consuming and tedious. In this paper, we introduce a registration algorithm that combines active contour segmentation together with mutual information. Our approach starts with a segmentation procedure. It is formed by a novel geometric active contour, which incorporates edge knowledge, namely Edgeflow, into active contour model. Two edgemap images filled with closed contours are obtained. After ruling out mismatched curves, we use mutual information (MI) as a similarity measure to register two edgemap images. Experimental results are provided to illustrate the performance of the proposed registration algorithm using both synthetic and multisensor images. Quantitative error analysis is also provided and several images are shown for subjective evaluation.     

Joint model of gradient magnitude and Gabor features via Spatio-Temporal slice

To form a high-performance video quality predictor, we developed a framework for full-reference (FR) video quality assessment that integrates Spatio-temporal slice analysis (STS) to create a high-performance predictor of video quality. However, both gradient and Gabor are spatial–temporal structural capturers used for the simultaneous extraction of both spatial and temporal features. In this paper, we proposed a novel VQA algorithm via a joint model of gradient magnitude and Gabor features (JMG) between the STS images of the reference videos and their distorted counterparts to assess the degradation of video quality effectively. Firstly, gradient magnitude and the Gabor filter were constructed to extract the spatiotemporal features of the video sequence. However, the two-feature model combined to predict the perceptual quality of frames. This new proposed VQA model is known as the horizontal and time STS (HT-JMG) model. To further investigate the influence of spatial dissimilarity, we combined the frame-by-frame spatial T-JMG(S) factor with the HT-JMG and propose another VQA model, called the time, horizontal, and vertical STS (THV-JMG) model. Finally, the results of the experiment showed that the proposed method has a strong correlation with subjective perception and is competitive with state-of-the-art full reference VQA models.     

Saturation-based adaptive inverse gradient interpolation for Bayerpattern images

A saturation-based adaptive gradient interpolation algorithm is developed for Bayer pattern images obtained in single-sensor digital cameras. The proposed algorithm uses the concept of pseudo-saturation for 2×2 blocks of the image. The algorithm adapts the scheme of interpolation based on the classification of the pseudo-saturation of Bayer pattern images. Compared with other algorithms, the proposed method preserves fine details, reduces false colours, and distortions around edges and sharp features of the colour image reconstructed from different kinds of Bayer pattern images     

Conjugate gradient method applied to inverse scattering problem

A new reconstruction algorithm for diffraction tomography is presented. The algorithm is based on the minimization of a functional which is defined as the norm of the discrepancy between the measured scattering amplitude and the calculated one for an estimated object function. By using the conjugate gradient method to minimize the functional, one can derive an iterative formula for getting the object function. Numerical results for some two-dimensional scatterers show that the algorithm is very effective in reconstructing refractive index distributions to which the first-order Born approximation can not be applied. In addition, the number of iterations is reduced by using a priori information about the outer boundary of the objects. Furthermore, the method is not so sensitive to the presence of noise in the scattered field data     

Automatic body segmentation with graph cut and self-adaptive initialization level set (SAILS)

In this paper, we propose an automatic human body segmentation system which mainly consists of human body detection and object segmentation. Firstly, an automatic human body detector is designed to provide hard constraints on the object and background for segmentation. And a coarse-to-fine segmentation strategy is employed to deal with the situation of partly detected object. Secondly, background contrast removal (BCR) and self-adaptive initialization level set (SAILS) are proposed to solve the tough segmentation problems of the high contrast at object boundary and/or similar colors existing in the object and background. Finally, an object updating scheme is proposed to detect and segment new object when it appears in the scene. Experimental results demonstrate that our body segmentation system works very well in the live video and standard sequences with complex background.     

Excursions of adaptive algorithms via the Poisson clumpingheuristic

The authors detail the application of the Poisson clumping heuristic (PCH) to the least mean square (LMS) adaptive algorithm and its signed variants. Under certain conditions on the input and disturbance statistics, the parameter estimate errors form a Markov process. The PCH asserts that large excursions of the parameter estimates occur in clumps, and that these clumps are distributed in a Poisson manner with parameter λ_b, where b is the maximum parameter error. Expressions are derived for each of the four algorithms in the scalar case, which allows calculation of λ _b in a relatively straightforward manner. These values are compared to simulations of the algorithms. Given that the results are asymptotic in b, the close agreement between simulated and theoretical values is striking, even for very modest b. The four algorithms are then compared in terms of λ_b. Some observations are made regarding the relative performance of the four variants. No single LMS variant always outperforms the others. Suggestions are made as to how this technique might be applied in the vector case, and a crucial monotonicity property is verified     

基于泊松分布的日盲紫外电晕检测

针对日盲紫外电晕探测噪声大、信号微弱等特点,本文提出了一种利用泊松分布的日盲紫外电晕检测方法.为了表征紫外电晕信号的时空域特性,根据光电探测原理,构建了基于泊松分布的紫外电晕目标检测模型.针对紫外图像序列,首先根据暗噪声统计模型进行有无信号的判断,然后利用所建立的紫外电晕目标检测模型得到紫外图像的泊松概率映射图,最后基于改进的最大类间方差法(Otsu)实现紫外电晕目标检测与提取.本文方法在多组紫外图像序列中进行测试与对比,实验证明该方法可以准确地检测出紫外图像中电晕目标,方法结构简单,检测精度高,且性能鲁棒.     

Automatic gradient threshold determination for edge detection

We describe a method to automatically find gradient thresholds to separate edge from nonedge pixels. A statistical model that is the weighted sum of two gamma densities corresponding to edge and nonedge pixels is used to identify a threshold. Results closely match human perceptual thresholds even under low signal-to-noise ratio (SNR) levels.     

Computation of the inverse Hessian matrix using conjugate gradient methods

A method is presented to compute the inverse of the Hessian matrix in an optimization problem using the conjugate gradient algorithm of Hestenes and Stiefel [1]. It is shown how this may be used to refine the solution by a Newton-Raphson iteration for both finite and infinite dimensional optimization problems.     

Intravital leukocyte detection using the gradient inverse coefficient of variation

The problem of identifying and counting rolling leukocytes within intravital microscopy is of both theoretical and practical interest. Currently, methods exist for tracking rolling leukocytes in vivo, but these methods rely on manual detection of the cells. In this paper we propose a technique for accurately detecting rolling leukocytes based on Bayesian classification. The classification depends on a feature score, the gradient inverse coefficient of variation (GICOV), which serves to discriminate rolling leukocytes from a cluttered environment. The leukocyte detection process consists of three sequential steps: the first step utilizes an ellipse matching algorithm to coarsely identify the leukocytes by finding the ellipses with a locally maximal GICOV. In the second step, starting from each of the ellipses found in the first step, a B-spline snake is evolved to refine the leukocytes boundaries by maximizing the associated GICOV score. The third and final step retains only the extracted contours that have a GICOV score above the analytically determined threshold. Experimental results using 327 rolling leukocytes were compared to those of human experts and currently used methods. The proposed GICOV method achieves 78.6% leukocyte detection accuracy with 13.1% false alarm rate.     

Enhancing inverse halftoning via coupled dictionary training

Inverse halftoning is a challenging problem in image processing. Traditionally, this operation is known to introduce visible distortions into reconstructed images. This paper presents a learning-based method that performs a quality enhancement procedure on images reconstructed using inverse halftoning algorithms. The proposed method is implemented using a coupled dictionary learning algorithm, which is based on a patchwise sparse representation. Specifically, the training is performed using image pairs composed by images restored using an inverse halftoning algorithm and their corresponding originals. The learning model, which is based on a sparse representation of these images, is used to construct two dictionaries. One of these dictionaries represents the original images and the other dictionary represents the distorted images. Using these dictionaries, the method generates images with a smaller number of distortions than what is produced by regular inverse halftone algorithms. Experimental results show that images generated by the proposed method have a high quality, with less chromatic aberrations, blur, and white noise distortions.     

改进的基于泊松模型的极大似然恢复算法

从高速流场扰动的失真图像中恢复出满足制导系统所需精度的目标图像是需要研究的课题.Holmes的基于泊松模型的极大似然校正算法存在平凡解的问题,针对此问题,提出在原算法的极大似然概率估计函数中增加适当的修正项来约束迭代校正算法向正确的方向收敛.讨论了修正项中参数的优化问题,提出按照使校正后图像的熵最小的优化准则来选取合适的修正参数.最后使用了实际风洞模拟实验中获得的气动光学效应降晰图像来测试改进算法的校正效果,同时给出原算法的校正结果以便比较.测试结果表明,改进算法有效消除了原算法的平凡解问题,校正效果较原算法有较为明显的提高.     

Automatic aircraft landing using interferometric inverse syntheticaperture radar imaging

This paper presents an interferometric processing of an aircraft's monostatic and bistatic inverse synthetic aperture radar (ISAR) signatures for automatic landing. The aircraft's squint angle in this ISAR imaging problem is near 90 degrees . We show that this extreme squint angle does not pose any problem for the ISAR Fourier-based (wavefront) reconstruction algorithm. In fact, the aircraft can be imaged accurately, and without any erroneous shifts in the cross-range domain, within the imposed theoretical resolution. Moreover, the algorithm is accurate enough such that one can utilize the phase of the ISAR monostatic and bistatic measurements for interferometric processing. The resultant interferometric ISAR image is used to detect undesirable rotations in the aircraft's orientation.     

Noise relations of inverse active networks and complementary networks

A direct relation is derived to calculate the noise of an inverse active network from the noise of its pair. A similar relation for complementary networks is also derived. Applications of these relations are given.     

Automatic analysis of two-port active microwave network

A new six-port ANA technique and its application in analysing an active microwave network are discussed. The reported isolated dual-six-port ANA (IDSPNA) provides a simple and powerful means to determine any two-port network scattering matrix. Computer simulation and experimental results for MIC amplifiers are presented.     

Robust online tracking via sparse gradient convolution networks

Convolution networks trained offline have recently exhibited promising performance in object tracking tasks. However, offline training is time-consuming and their performance heavily rely on the category of auxiliary training sets. In this paper, we propose a sparse gradient convolution network without pretraining for object tracking. This approach combines shallow convolutional networks and traditional methods (gradient features and sparse representations) to avoid the offline training. In the first frame, we utilize the sparse representation method to learn a series of gradient-based local patches served as fixed filters, and they are used to convolving the input image in the subsequent frames to encode local structural information. Then, we stack all the local structure features to construct global spatial structure features, and the inner geometric layout information is preserved. Moreover, sparse coding and online updating are used to overcome issues related to target appearance variations. Qualitative and quantitative evaluations based on a challenging benchmark dataset demonstrate the effectiveness of the proposed algorithm against several state-of-the-art tracking methods.     

Image segmentation and analysis via multiscale gradient watershedhierarchies

Multiscale image analysis has been used successfully in a number of applications to classify image features according to their relative scales. As a consequence, much has been learned about the scale-space behavior of intensity extrema, edges, intensity ridges, and grey-level blobs. We investigate the multiscale behavior of gradient watershed regions. These regions are defined in terms of the gradient properties of the gradient magnitude of the original image. Boundaries of gradient watershed regions correspond to the edges of objects in an image. Multiscale analysis of intensity minima in the gradient magnitude image provides a mechanism for imposing a scale-based hierarchy on the watersheds associated with these minima. This hierarchy can be used to label watershed boundaries according to their scale. This provides valuable insight into the multiscale properties of edges in an image without following these curves through scale-space. In addition, the gradient watershed region hierarchy can be used for automatic or interactive image segmentation. By selecting subtrees of the region hierarchy, visually sensible objects in an image can be easily constructed.