We develop a general framework to simultaneously exploit texture and shape characterization in multiscale image segmentation. By posing multiscale segmentation as a model selection problem, we invoke the powerful fram...
详细信息
ISBN:
(纸本)0819437646
We develop a general framework to simultaneously exploit texture and shape characterization in multiscale image segmentation. By posing multiscale segmentation as a model selection problem, we invoke the powerful framework offered by minimum description length (MDL). This framework dictates that multiscale segmentation comprises multiscale texture characterization and multiscale shape coding. Analysis of current multiscale maximum a posteriori (MAP) segmentation algorithms reveals that these algorithms implicitly use a shape coder with the aim to estimate the optimal MDL solution, but find only an approximate solution. Towards achieving better segmentation estimates, we first propose a shape coding algorithm based on zero-trees which is well-suited to represent images with large homogeneous regions. For this coder, we design an efficient tree-based algorithm using dynamic programming that attains the optimal MDL segmentation estimate. To incorporate arbitrary shape coding techniques into segmentation, we design an iterative algorithm that uses dynamic programming for each iteration. Though the iterative algorithm is not guaranteed to attain exactly optimal estimates, it more effectively captures the prior set by the shape coder. Experiments demonstrate that the proposed algorithms yield excellent segmentation results on both synthetic and red world data examples.
wavelet representations of images are increasingly important as more imageprocessing functions are shown to be advantageously executed in the wavelet domain. images may be inverse halftoned, compressed, denoised, and...
详细信息
ISBN:
(纸本)0819448087
wavelet representations of images are increasingly important as more imageprocessing functions are shown to be advantageously executed in the wavelet domain. images may be inverse halftoned, compressed, denoised, and enhanced in the wavelet domain. In conjunction with other waveletprocessing, it would be efficient to halftone directly from the wavelet domain. In this paper we demonstrate how to perform error diffusion in the wavelet domain. The wavelet coefficients are modified by a normalization factor and re-arranged. Then, traditional feed-forward raster scan error diffusion is performed and quality halftones are shown to result. Error diffusing in the wavelet domain is noted to be non-causal with respect to the pixels, and thus the method is not reproducible by feed-forward raster scan error diffusion of pixels. It is shown that the wavelet halftones preserve the average value of the input for constant patches. The resulting halftones may appear smoother in smooth regions and sharper at edges than the corresponding pixel-domain halftones. Disadvantages may include a greater susceptibility to moire and false contouring. Error diffusion is a two-dimensional sigma-delta modulation, and the ideas presented may also be useful for one-dimensional sigma-delta modulation applications.
We develop a new class of non-Gaussian multiscale stochastic processes defined by random cascades on trees of wavelet or other multiresolution coefficients. These cascades reproduce a rich semi-parametric class of ran...
详细信息
ISBN:
(纸本)0819437646
We develop a new class of non-Gaussian multiscale stochastic processes defined by random cascades on trees of wavelet or other multiresolution coefficients. These cascades reproduce a rich semi-parametric class of random variables known as Gaussian scale mixtures. We demonstrate that this model class can accurately capture the remarkably regular and non-Gaussian features of natural images in a parsimonious fashion, involving only a small set of parameters. In addition, this model structure leads to efficient algorithms for imageprocessing. In particular, we develop a Newton-like algorithm for MAP estimation that exploits very fast algorithms for linear-Gaussian estimation on trees, and hence is efficient. On the basis of this MAP estimator, we develop and illustrate a denoising technique that is based on a global prior model, and preserves the structure of natural images (e.g., edges).
Most digital signalprocessing methods have an underlying assumption of regularly-spaced data samples. However, many real-world data collection techniques generate data sets which are not sampled at evenly-spaced inte...
详细信息
ISBN:
(纸本)0819422134
Most digital signalprocessing methods have an underlying assumption of regularly-spaced data samples. However, many real-world data collection techniques generate data sets which are not sampled at evenly-spaced intervals, or which may have significant data dropout problems. Therefore, a method of interpolation is needed to model the signal on an even grid of arbitrary granularity. We propose the interpolation of nonuniformly sampled fields using a least- square fit of the data to a wavelet basis in a multiresolution setting.
14-bit or 16-bit pixel depth high dynamic-range images are acquired from visible band cameras and from infrared imaging devices which are more widely used nowadays. Usually, linear mapping is used to display these ima...
详细信息
ISBN:
(纸本)9781467355636;9781467355629
14-bit or 16-bit pixel depth high dynamic-range images are acquired from visible band cameras and from infrared imaging devices which are more widely used nowadays. Usually, linear mapping is used to display these images to operators. However, results of the researches done to map images into 0-255 range in recent years show that different techniques result in major differences at image perception and detail visibility. Successful compression of image dynamic range increases the operator awareness for surveillance systems and ensure more effective display of scene details to user. Besides, dynamic-range compression techniques effect the enhancement of the success rate of image target detection and tracking techniques. In this work, scene components are analyzed using wavelet coefficients and intensity distribution of scene components are extracted. Extracted intensity distribution is used to display scene components effectively.
In this paper, we propose an image restoration algorithm based on state-of-the-art wavelet domain statistical models. We present an efficient method to estimate the model parameters from the observations, and solve th...
详细信息
ISBN:
(纸本)0819441929
In this paper, we propose an image restoration algorithm based on state-of-the-art wavelet domain statistical models. We present an efficient method to estimate the model parameters from the observations, and solve the restoration problem in orthonormal and translation-invariant (TI) wavelet domains. Substantial improvements over previous wavelet-based restoration methods are obtained. The use of a TI wavelet transform further enhances the restoration performance. We study the improvement from the viewpoint of Bayesian estimation theory and show that replacing an estimator with its TI version will reduce the expected risk if the signal and the degradation model are stationary.
This paper investigates the relationship between the traditional wavelet (or matched filter) detector and the estimator correlator (EC) detector formulated in the wavelet domain. The EC detector is actually a weighted...
详细信息
ISBN:
(纸本)0819425915
This paper investigates the relationship between the traditional wavelet (or matched filter) detector and the estimator correlator (EC) detector formulated in the wavelet domain. The EC detector is actually a weighted wavelet detector, weighted by the scattering function that describes the medium and/or model. The wavelet detector is the optimum detector for point objects but it does not incorporate knowledge of the scattering environment. However, when imaging distributed objects, it is advantageous to take a priori information into account. The EC incorporates this information as a weight on the waveletimage and formulates an estimated spreading function which essentially achieves recombination of highlights and multipath energy. It can be shown that the EC reduces to the the wavelet detector when a point object is being imaged.
The objective measurement of blockiness plays an important role in many applications, such as the quality assessment of an image, and the design of image and video coding system. However, most of the existing no-refer...
详细信息
ISBN:
(纸本)9781479903566
The objective measurement of blockiness plays an important role in many applications, such as the quality assessment of an image, and the design of image and video coding system. However, most of the existing no-reference blockiness metrics do not consider important influences of grid distortion of an image on the performance of the metric. In this paper, we propose a new blockiness metric, which is robust to grid distortion, based on the marginal distribution of local wavelet coefficients and saliency information. Experiments for several public image databases showed that the proposed metric provides consistent correlations with subjective blockiness scores and outperforms other existing no-reference blockiness metrics.
This paper introduces applications of wavelet analysis in mechanical signalprocessing based on the therory wavelet analysis and wavelet recognition. It can effectively distinguish different running states of rolling ...
详细信息
ISBN:
(纸本)9780819489326
This paper introduces applications of wavelet analysis in mechanical signalprocessing based on the therory wavelet analysis and wavelet recognition. It can effectively distinguish different running states of rolling bearing through picking up character of scale-energy. It is one of the important characteristics of signal singularity and irregular mutation that it includes important information. Analysis singularity and the site of singularity using wavelet transform is more effective because wavelet transform has space localization property. In wavelet analysis, the low frequency part of signal represents the development trend of the signal. It corresponds to the maximum scale wavelet coefficients. With the increase of scale, the resolution of time, the development trend of the signal can be more apparent, so wavelet analysis on display the development trend of signal is very useful. Along with the computer hardware technology and specially the DSP technology development, it will provide powerful safeguard for wavelet analysis using project.
暂无评论