基于Log-Gabor滤波的指纹图像增强

自动指纹识别系统的性能在很大程度上依赖于指纹图像的采集效果。指纹图像增强用于改善原始指纹图像的质量,以保证指纹特征提取的准确性,是十分必要的。Gabor滤波是比较有效的指纹图像增强方法,但是仍存在一定的局限性。与传统的Gabor滤波器相比,Log-Gabor滤波器可以在取得最佳空间定位的同时具有更宽广的频带,有利于改善指纹图像的滤波效果。本文提出采用Log-Gahor滤波器来实现指纹图像增强,先利用加窗傅里叶变换来提取指纹图像的局部频谱信息,再在频域进行滤波。在详述了滤波器的设计方法之后,给出具体的图像滤波方案,并与传统方法作了比较。实验结果表明,所提出的算法能有效改善指纹图像的质量和提高指纹识别的可靠性。     

Scale-space properties of quadratic feature detectors

We consider the scale-space properties of quadratic feature detectors and, in particular, investigate whether, like linear detectors, they permit a scale selection scheme with the “causality property”, which guarantees that features are never created as the scale is coarsened. We concentrate on the design of one dimensional detectors with two constituent filters, with the scale selection implemented as convolution and a scaling function. We consider two special cases of interest: the constituent filter pairs related by the Hilbert transform, and by the first spatial derivative. We show that, under reasonable assumptions, Hilbert-pair quadratic detectors cannot have the causality property. In the case of derivative-pair detectors, we describe a family of scaling functions related to fractional derivatives of the Gaussian that are necessary and sufficient for causality. In addition, we report experiments that show the effects of these properties in practice. We thus demonstrate that at least one class of quadratic feature detectors has the same desirable scaling property as the more familiar detectors based on linear filtering     

Textural filters based on the texture spectrum

Conventional digital filtering techniques, based on classical Fourier analysis (that is, low-pass, high-pass and band-pass), are widely used in digital image processing. Unsatisfaction may be encountered when applying these filters to texture analysis of images, where one needs some specific spatial filters which are able to transform an image in the sense of texture rather than the spectral properties. Such textural filters can be designed in the texture spectrum domain and they are of interest for texture analysis. An example is given in this paper, and has been applied to four of Brodatz's natural images. The result shows a promising potential of the texture spectrum for designing textural filters.     

图像引导滤波的局部多尺度Retinex算法

Retinex算法是一种用于消除由光照变化给图像所带来的负面影响的图像增强算法。该算法的求解通常需要基于入射分量分段光滑的假设,利用正则化的方法迭代求解,计算效率低。文中基于一项最近提出的研究——"图像引导滤波",提出一种非迭代的Retinex算法框架。基于反射分量也满足分段光滑的假设,采用两次图像引导滤波克服了图像噪声所带来的影响。然后在基于小波变换域图像融合策略的基础上,提出基于图像引导滤波的多尺度Retinex算法,实现图像细节增强与颜色保真之间的平衡。实验结果表明,与各种算法相比,该算法在克服噪声、细节增强和颜色保真方面能够取得更好的效果。     

SIFER: Scale-Invariant Feature Detector with Error Resilience

We present a new method to extract scale-invariant features from an image by using a Cosine Modulated Gaussian (CM-Gaussian) filter. Its balanced scale-space atom with minimal spread in scale and space leads to an outstanding scale-invariant feature detection quality, albeit at reduced planar rotational invariance. Both sharp and distributed features like corners and blobs are reliably detected, irrespective of various image artifacts and camera parameter variations, except for planar rotation. The CM-Gaussian filters are approximated with the sum of exponentials as a single, fixed-length filter and equal approximation error over all scales, providing constant-time, low-cost image filtering implementations. The approximation error of the corresponding digital signal processing is below the noise threshold. It is scalable with the filter order, providing many quality-complexity trade-off working points. We validate the efficiency of the proposed feature detection algorithm on image registration applications over a wide range of testbench conditions.     

Robust scale invariant target detection using the scale-space theory and optimization for IRST

This paper presents a new small target detection method using scale invariant feature. Detecting small targets whose sizes are varying is very important to automatic target detection in infrared search and track (IRST). The conventional spatial filtering methods with fixed sized kernel show limited target detection performance for incoming targets. The scale invariant target detection can be defined as searching for maxima in the 3D (x, y, and scale) representation of an image with the Laplacian function. The scale invariant feature can detect different sizes of targets robustly. Experimental results with real FLIR images show higher detection rate and lower false alarm rate than conventional methods. Furthermore, the proposed method shows very low false alarms in scan-based IR images than conventional filters.     

Fusion of multispectral and panchromatic satellite sensor imagery based on tailored filtering in the Fourier domain

A new methodology for fusing satellite sensor imagery, based on tailored filtering in the Fourier domain is proposed. Finite‐duration Impulse Response (FIR) filters have been designed through an objective criterion, which depends on source image characteristics only. The designed filters allow a weighted fusion of the information contained in a fine spatial resolution image (PAN) and in a multispectral image (MULTI), respectively, establishing a trade‐off between spatial and spectral quality of the resulting fused image. This new technique has been tested with Landsat Enhanced Thematic Mapper Plus (ETM+) imagery. Spatial and spectral quality of the fused images was compared with the results provided by Mallat's Wavelet algorithm. The images fused by the proposed method were characterized by a spatial resolution very close to the PAN image, and by the spectral resolution of the MULTI image.     

Speckle reduction for the forest mapping analysis of multi-temporal Radarsat-1 images

As the number of satellite-borne synthetic aperture radar (SAR) systems increases, both the availability and the length of multi-temporal (MT) sequences of SAR images have also increased. Previous research on MT SAR sequences suggests that they increase the classification accuracy for all applications over single date images. Yet the presence of speckle noise remains a problem and all images in the sequence must be speckle filtered before acceptable classification accuracy can be attained. Several speckle filters designed specifically for MT sequences have been reported in the literature. Filtering in the spatial domain, as is usually done, reduces the effective spatial resolution of the filtered image. MT speckle filters operate in both the spatial and temporal dimensions, thus the reduction in resolution is not likely to be as severe (although a comparison between MT and spatial filters has not been reported). While this advantage may be useful when extracting spatial features from the image sequence, it is not quite as apparent for classification applications. This research explores the relative performance of spatial and MT speckle filtering for a particular classification application: mapping boreal forest types. We report filter performance using the radiometric resolution as measured by the equivalent number of looks (NL), and classification performance as measured by the classification accuracy. We chose representative spatial and MT filters and found that spatial speckle filters offer the advantage of higher radiometric resolution and higher classification accuracy with lower algorithm complexity. Thus, we confirm that MT filtering offers no advantage for classification applications; spatial speckle filters yield higher overall performance.     

A state variable approach to digital image processing

Digital filtering of images is considered. A recursive filter design for two-dimensional separable filters is discussed which is based on state variable methods. Advantages of this design are guaranteed stability, and shorter word length than filters designed by bilinear mapping techniques, greater computational efficiency and easier design as compared with nonrecursive filters. Results are shown for filtering high frequency additive noise from a low spatial frequency image and for filtering a low frequency multiplicative noise from a high spatial frequency image.     

Scale‐adaptive Structure‐preserving Texture Filtering

This paper proposes a scale‐adaptive filtering method to improve the performance of structure‐preserving texture filtering for image smoothing. With classical texture filters, it usually is challenging to smooth texture at multiple scales while preserving salient structures in an image. We address this issue in the concept of adaptive bilateral filtering, where the scales of Gaussian range kernels are allowed to vary from pixel to pixel. Based on direction‐wise statistics, our method distinguishes texture from structure effectively, identifies appropriate scope around a pixel to be smoothed and thus infers an optimal smoothing scale for it. Filtering an image with varying‐scale kernels, the image is smoothed according to the distribution of texture adaptively. With commendable experimental results, we show that, needing less iterations, our proposed scheme boosts texture filtering performance in terms of preserving the geometric structures of multiple scales even after aggressive smoothing of the original image.     

Making optical Fourier filtering work

During the work by many scientists over the last 50 years, we have solved many of the problems that seemed to be holding the field back: System layout, lenses specialized for this purpose, suitable detector arrays, SLMs for input and for filtering, designs for filters, online evolution of filters, and so forth. Yet it is hard to find this technology in use anywhere. Something is clearly wrong, and we know one major problem that has never been solved and, in fact, is seldom acknowledged in print: Only very simple problems are linearly discriminable, so they cannot be discriminated well by a linear discriminant. But, the wonderful property of correlating all parts of the input pattern in parallel is only possible by Fourier methods. We see no way whatever for overcoming the contradictory goals of space or time invariant operations and need of implementing the most powerful nonlinear discriminants (clearly this can only be done with nonlinear discriminants). Those two contradictory goals seem to condemn Fourier filtering (electronic or optical) to being interesting but not very helpful. As this is a paper on a new kind of pattern recognition and it is not limited to Fourier optics, it is important to provide some useful concepts in Pattern Recognition. Our discussion is simple and brief. It has no proofs but it does offer the reader some of the concepts that help in understanding the new aspects of the material presented here. This work will not propose or develop improvements on any of the configurations and components. This is a study of the way Fourier filtering can have it both ways. Prior analyses are still valid, but, at a higher level, we can have it both ways. This review is aimed at allowing both the space or time invariant filtering while implementing an extremely powerful linear discriminants. Aspects of this have been published but in a regrettably short form and with very little of the background on pattern recognition. Here, we begin with a brief introduction to pattern recognition as will be needed to make this review finite. After the way to achieve powerful target discrimination and target location is shown, I describe how to apply it to optical Fourier filtering for pattern recognition. Some situations seem more likely to use Optical Fourier pattern recognition than others. We explore some of those.     

Efficient synthesis of gaussian filters by cascaded uniform filters

Gaussian filtering is an important tool in image processing and computer vision. In this paper we discuss the background of Gaussian filtering and look at some methods for implementing it. Consideration of the central limit theorem suggests using a cascade of ``simple' filters as a means of computing Gaussian filters. Among ``simple' filters, uniform-coefficient finite-impulse-response digital filters are especially economical to implement. The idea of cascaded uniform filters has been around for a while [13], [16]. We show that this method is economical to implement, has good filtering characteristics, and is appropriate for hardware implementation. We point out an equivalence to one of Burt's methods [1], [3] under certain circumstances. As an extension, we describe an approach to implementing a Gaussian Pyramid which requires approximately two addition operations per pixel, per level, per dimension. We examine tradeoffs in choosing an algorithm for Gaussian filtering, and finally discuss an implementation.     

Adaptive scale filtering: a general method for obtaining shape fromtexture

Introduces adaptive scale filtering, a general method for deriving shape from texture under perspective projection without recourse to prior segmentation of the image into geometric texture elements (texels), and without thresholding of filtered images. If texels on a given surface can be identified in an image then the orientation of that surface can be obtained. However, there is no general characterization of texels for arbitrary textures. Furthermore, even if the size and shape of texels on the surface is invariant with regard to position, perspective projection ensures that the size and shape of the corresponding image texels vary by orders of magnitude. Commencing with an initial set F_O of identical image filters, adaptive scale filtering iteratively derives a set F_N which contains a unique filter for each image position. Each element of F_N is tuned to the three-dimensional structure of the surface; that is, all image filters in F_N back-project to an identical shape and size on the surface. Thus image texels of various sizes, but associated with a single spatial scale on the surface, can be identified in different parts of the image. When combined with conventional shape from texture methods, edges derived using F_N provide accurate estimates of surface orientation. Results for planar surfaces are presented     

Separability of Polyhedra for Optimal Filtering of Spatial and Constraint Data

The filtering method considered in this paper is based on approximation of a spatial object in d-dimensional space by the minimal convex polyhedron that encloses the object and whose facets are normal to preselected axes. These axes are not necessarily the standard coordinate axes; furthermore, their number is not determined by the dimension of the space. We optimize filtering by selecting optimal such axes based on a preprocessing analysis of stored objects or a sample thereof. The number of axes selected represents a trade-off between access time and storage overhead, as more axes usually lead to better filtering but require more overhead to store the associated access structures. We address the problem of minimizing the number of axes required to achieve a predefined quality of filtering and the reverse problem of optimizing the quality of filtering when the number of axes is fixed. In both cases we also show how to find an optimal collection of axes. To solve these problems, we introduce and study the key notion of separability classification, which is a general tool potentially useful in many applications of a computational geometry flavor. The approach is best suited to applications in which the spatial data is relatively static, some directions are more dominant than others, and the dimension of the space is not high; maps are a prime example.     

Filtering method for SAR interferograms with strong noise

We develop a filtering method for SAR interferograms with strong noise based on the concept of the Vondrák filter and the commonly used Goldstein filter. The one‐dimensional Vondrák filter is first extended to two dimensions, and then utilized to smooth the Fourier spectra of overlapped interferogram patches. The smoothed spectra are then back‐transformed into the interferogram space. Experimental results with simulated and real datasets show that the proposed filter can eliminate up to 70% of noise in a SAR interferogram. The results also show a 15% to 35% improvement over some of the existing filters when the noise in an interferogram is strong, and the stronger the noise in an interferogram, the greater is the improvement.     

Filtering procedures for flow in heterogeneous porous media: numerical results

This paper focuses on heterogeneous soil permeabilities and on the impact their resolution has on the solution of the piezometric head equation. The method of coarse graining is proposed in order to filter the piezometric head equation on arbitrary support scales: Large scale fluctuations of the permeabilities are resolved, whereas small scale fluctuations are smoothed by a spatial filtering procedure. The filtering procedure is performed in Fourier space with the aid of a low-frequency cut-off function. In the filtered equations, the impact of the small scale variability is modeled by scale dependent effective permeabilities which are determined by additional differential equations. The additional differential equations are equivalent to the piezometric head equation on cells which are solved numerically by using the software toolbox UG. The numerical results are compared with the theoretical results derived in Attinger 2001.     

分数阶小波变换

小波变换是对信号时域-频域(Fourier域)的多分辨率分析,也可看作是一种Fourier域伸缩带通滤波.分数阶Fourier变换是对传统Fourier变换的推广,对信号分析处理有更大的灵活性,为了将多分辨率分析理论推广到时域-广义频域(分数阶Fourier域),提出了一种分数阶小波变换,分析了分数阶小波变换在广义频域伸缩带通滤波特性,分析信号时的时域-广义频域平面的多分辨率分析网格划分.分数阶小波变换是传统小波变换的推广,在对原小波变换核作一定改动后增加了小波变换对信号处理的灵活性.可以看到,将分数阶小波变换的变换角度取为π/2,便得到与传统小波变换多分辨率分析理论完全一致的结果.理论分析和计算机仿真表明了所提理论的正确性和有效性.     

结合双边滤波和多帧均值滤波的图像降噪

针对单幅图像空间降噪中存在的问题,提出一种利用同一场景的多帧图像进行噪声抑制的自适应空域滤波方法。首先对每一帧用双边滤波器进行滤波,以在平滑噪声的同时,保护图像边缘;然后对每一像素进行帧间平均值滤波,以利用多幅图像的信息进一步滤除噪声,并减轻双边滤波后的卡通效果。提出的算法在两组多帧图像上进行了测试。实验显示,与单帧双边滤波和简单的多图像平均方法相比,该方法的降噪性能有所提高。     

Filtering of satellite images in geological lineament analyses: An application to a fault zone in Central Turkey

Lineament extraction and analysis is one of the routines in mapping large areas using remotely-sensed data, most of which is the satellite images. In this study, we aimed to test different lineament extraction techniques including single band, multiband enhancements and spatial domain filtering techniques. A fast algorithm has been developed for time and cost limited surveys in an area with known dominant and/or any selected orientation of lineaments. During the study for single band analysis, histogram equalization and stretching, for multiband, Principal Components Analysis (PCA) and for spatial domain filtering Prewitt and Sobel filters are utilized. Furthermore, we have developed a new algorithm which consists of a combination of large smoothing filters and gradient filters, in order to get rid of the artificial lineaments which are out of interest and to determine discontinuous and/or closely spaced regional lineaments. The results are as follows; the number of lineaments and their total lengths are 218 285.3 km using the single band; 255 343.9 km in multiband enhancements, and 347644.9 km in the combination of spatial domain filtering including our algorithm. The 59.17 per cent increase in the number of lineaments and the 126.04 per cent increase in the total lengths indicate that a combination of spatial domain filters is the most cost-time efficient algorithm in lineament analyses.     

A fuzzy filter for the removal of random impulse noise in image sequences

In this paper a new fuzzy filter for the removal of random impulse noise in digital grayscale image sequences is presented. The filter consists of different noise detection and filtering steps, in which the fuzzy set theory is used. This noise detection is based both on spatial and on temporal information and has the aim to prevent the filtering of noise free image pixels. The filtering of the detected noisy pixels is finally performed in a motion compensated way. Experimental results show that our method outperforms other state-of-the-art filters in terms of the peak-signal-to-noise ratio as well as visual quality.