A.R. Rao, A Taxonomy for Texture Description and Identification, Springer Verlag 1990.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....as we show later in this paper, the proposed measure can also be related to the Bayes error. There is a large body of image processing literature which deals with quantifying the texture of an image (in image processing, pixel intensities may be viewed as serving the role of the class label) [5]. Here we choose one of the most common statistical approach for the characterization of texture the co occurrence matrix [6, 5] Within the context of image processing, the co occurrence matrix represents the second order joint conditional density function f(i, j d, #) i.e. the probability ....

.... processing literature which deals with quantifying the texture of an image (in image processing, pixel intensities may be viewed as serving the role of the class label) 5] Here we choose one of the most common statistical approach for the characterization of texture the co occurrence matrix [6, 5]. Within the context of image processing, the co occurrence matrix represents the second order joint conditional density function f(i, j d, #) i.e. the probability of going from gray level i to gray level j within a distance d along a direction #. While the direction is relevant within the ....

[Article contains additional citation context not shown here]

A. R. Rao, A Taxonomy for Texture Description and Identification, Springer-Verlag, New York, New York, 1990.

....which can be automatically and reliably extracted from the fingerprint and whose extraction will degrade gracefully with deterioration in the quality of the fingerprints. 3. 1 Introduction The smooth flow pattern of ridges and valleys in a fingerprint can be viewed as an oriented texture field [28] (see Figure 3.1) The image intensity surface in fingerprint images is comprised of ridges whose directions vary continuously, which constitutes an oriented texture. Most textured images contain a limited range of spatial frequencies, and mutually distinct textures di#er significantly in their ....

....coherence at each pixel in the image. Fingerprints can be represented matched by using quantitative measures associated with the flow pattern (oriented texture) as features. Analysis and modeling of oriented textures is an important research problem with a wide variety of practical applications [28]. Previous attempts at describing oriented textures have used either exclusively local or predominantly global features. Examples of local representations include Poincare indices, winding numbers, and information related to singularities and anomalies. The primary limitation of the local ....

[Article contains additional citation context not shown here]

A. R. Rao, A Taxonomy for Texture Description and Identification, SpringerVerlag, New York, 1990.

....[14, 23, 15] and the rotation angle of a cross. In the Cartesian coordinate system, the linear symmetry theory is the mathematical basis of creating dense complex vector (tensor) fields representing local orientation and certainties of straight lines. Some applications are illustrated by texture [30, 6] and finger print analysis as well as image enhancement, 18] However, when the method is applied to other (harmonic) coordinate systems efficient implementation becomes a non trivial issue. For illustration, in Section 3, we will elaborate on one such coordinate system that is not Cartesian but ....

A. R. Rao. A taxonomy for texture description and identification. Springer, 1990.

....Lenders et al. 138] have proposed a 1 bit systolic processor for binary morphology. 4.4 Application of generalized convolution In this section, two applications of generalized convolution are described. For computations of dominant ridge directions in a fingerprint, the orientation field [182] model is adopted. In order to remove spiky growths in a fingerprint skeleton image, a morphological filter has been described in [184] For removing repetitive background in document images, Liang et al. 142] proposed a morphological approach. These three applications will be used to ....

A. R. Rao. A Taxonomy for Texture Description and Identification. SpringerVerlag, New York, 1990.

....directional derivatives, Kass and Witkin [12] came to a solution that turned out to be equivalent to the tensor method. Searching for a general description of local orientation Knutsson [13] concluded that local structure in an n dimensional space can be represented by a symmetric nn tensor. Rao [19] used a similar tensor representation for 2D texture analysis. 2.2 Total Least Squares Optimization The orientation of gray value structures can mathematically be formulated as a total least squares optimization problem [3] The scalar product between a vector r , representing the orientation ....

A. R. Rao, 1990. A Taxonomy for Texture Description and Identification, Springer, New York 3

....A texture feature is used to characterize texture in an image. The result of the characterization is a measure of how well the texture can be identified by that feature. This measure is then used by a region partitioning algorithm or other texture discrimination algorithm to segment an image. Rao [38] developed a taxonomy for the description of different types of textures which could effectively classify a set of synthetic images (the Brodatz texture images) This taxonomy provides a standardized set of descriptors which can be used for qualitative descriptions of textures. Thus, a means is ....

A.R. Rao. A Taxonomy for Texture Description and Identification. SpringerVerlag, 1990.

....investigated for texture related applications. However, what properties are perceptually meaningful is seldom considered. Tamura et al. 17] and Amadasun et al. 1] had earlierly proposed many visually relevant texture features. Afterwards, based on intuition and computational considerations, Rao [13] provided a texture taxonomy as strongly ordered, weakly ordered, and disordered textures. Followed, Rao et al. 14] 15] identified relevant dimensions of textures and found that the most perceptual texture features are periodicality , directionality , and complexity from the psychological ....

A. R. Rao, "A Taxonomy for Texture Description and Identification", Springer-Verlag, 1990.

....Field Fingerprint Region Thinning Localization Extraction Minutia Extraction Ridge Fingerprint Image Minutiae Figure 7: Flowchart of the minutia extraction algorithm. 2. 1 Estimation of Orientation Field A number of methods have been proposed to estimate the orientation field of flow like patterns [17]. In our system, a new hierarchical implementation of Rao s algorithm [17] is used. Rao s algorithm consists of the following main steps: 1. Divide the input fingerprint image into blocks of size W Theta W . 2. Compute the gradients G x and G y at each pixel in each block. 3. Estimate the local ....

....Ridge Fingerprint Image Minutiae Figure 7: Flowchart of the minutia extraction algorithm. 2. 1 Estimation of Orientation Field A number of methods have been proposed to estimate the orientation field of flow like patterns [17] In our system, a new hierarchical implementation of Rao s algorithm [17] is used. Rao s algorithm consists of the following main steps: 1. Divide the input fingerprint image into blocks of size W Theta W . 2. Compute the gradients G x and G y at each pixel in each block. 3. Estimate the local orientation of each block using the following formula: o = 1 2 tan ....

A. Ravishankar Rao, A Taxonomy for Texture Description and Identification, SpringerVerlag, New York, 1990.

....flowchart is depicted in Figure 2. Orientation Estimation Field Fingerprint Region Thinning Localization Extraction Minutia Extraction Ridge Minutiae Figure 2. Flowchart of the minutia extraction algorithm 2.1. Estimation of Orientation Field A new hierarchical implementation of Rao s algorithm [7] is used to estimate the orientation field of an input fingerprint image. It consists of the following two main steps: 1. Estimate the local orientation at each pixel using following formula: o = 1 2 tan Gamma1 ( P W i=1 P W j=1 2G x (i; j)G y (i; j) P W i=1 P W j=1 (G 2 x (i; j) ....

A. Ravishankar Rao, A Taxonomy for Texture Description and Identification, Springer-Verlag, New York, 1990.

....Such a monitor could be linked to control loops. The method we use takes the Prewitt operator as described in [4] for greylevel images and obtains two variables. These are the local gradient magnitude and direction for the image. The concept of local coherence is taken from texture analysis [5] and provides us with an estimate of local anisotropy. 2 Variability of Local Grammage The variability of areal density, or grammage, in a 2 dimensional image is measured by the local average g(x; y) in a finite zone about the point (x; y) In order to infer some of the properties for the local ....

....properties of the sample as a whole end up varying with orientation. In this section we discuss how to detect and measure local anisotropy by finding maxima of the projections of the local gradients onto axes with particular orientations. These orientations are called local dominant orientations [5], and we discuss next an approach to detect them. 3.1 Estimating Local Dominant Orientations Local anisotropy is detected through local dominant orientations. These are orientations in which the grammage tends to vary most slowly due to an underlying anisotropic process, and consequently present ....

[Article contains additional citation context not shown here]

A. R. Rao. A Taxonomy for Texture Description and Identification. Ed. Springer-Verlag, Berlin, Germany, 1990.

....dpi. 2.1 Orientation Field Estimation The orientation field of a fingerprint image represents the intrinsic nature of the fingerprint image. It plays a very important role in fingerprint image analysis. A number of methods have been proposed to estimate the orientation field of fingerprint images [40, 56, 38]. In our system, a new hierarchical implementation of the algorithm proposed in [56] is used (Figure 8) With this algorithm, a fairly smooth orientation field estimate can be obtained. Figure 9 shows the orientation field of a fingerprint image estimated with our hierarchical algorithm. ....

....the intrinsic nature of the fingerprint image. It plays a very important role in fingerprint image analysis. A number of methods have been proposed to estimate the orientation field of fingerprint images [40, 56, 38] In our system, a new hierarchical implementation of the algorithm proposed in [56] is used (Figure 8) With this algorithm, a fairly smooth orientation field estimate can be obtained. Figure 9 shows the orientation field of a fingerprint image estimated with our hierarchical algorithm. Orientation Ridge Extraction Estimation Locator Fingerprint Thinning Minutia Extraction ....

[Article contains additional citation context not shown here]

A. R. Rao, A Taxonomy for Texture Description and Identification, Springer-Verlag, New York, 1990.

....texture analysis is an easy task, however. To develop a texture analysis system capable of automatic description and recognition of texture patterns is a difficult task, because of lack of mathematical models due to the complex nature of texture. The main tasks involved in texture analysis are [Rao, 1990]: Detecting and defining texture for identification and description of texture patterns. Image classification for image segmentation by using texture. Shape from texture for recovering information about surface orientation, shape and depth. This study addresses an adaptive method ....

....subdividing. 4.3 Determination of the Parameters The dominant orientation and spatial frequency are computed with the textons for each region which is defined by the progressive subdividing of the image. Determination of the Gabor filter size is discussed. 4.3. 1 Orientation In this study, Rao s [1990] inverse arctangent method is applied to compute the dominant local orientation of the textons for each region (i.e. image block from progressive subdividing) The underlying assumption is that the texture has only one dominant local orientation for each region. ....

[Article contains additional citation context not shown here]

Rao, A., 1990. A Taxonomy for Texture Description and Identification. Springer-Verlag, New York, Inc.

....of ridge directions in nonoverlapping windows in the fingerprint image. The predominant direction is computed as an optimal estimate of the direction vectors at each pixel in the window. Fingerprint images can be considered as an oriented texture pattern. As per the taxonomy described in [34], fingerprints can be classified as a weakly ordered texture. The orientation field [34] is used to compute the optimal dominant ridge direction in each 16 Theta 16 window or block. The following steps are involved in the computation of the orientation field for each window. 1. Compute the ....

....direction is computed as an optimal estimate of the direction vectors at each pixel in the window. Fingerprint images can be considered as an oriented texture pattern. As per the taxonomy described in [34] fingerprints can be classified as a weakly ordered texture. The orientation field [34] is used to compute the optimal dominant ridge direction in each 16 Theta 16 window or block. The following steps are involved in the computation of the orientation field for each window. 1. Compute the gradient of the smoothed block. Let G x (i; j) and G y (i; j) be the gradient magnitude in x ....

[Article contains additional citation context not shown here]

A. R. Rao. A Taxonomy for Texture Description and Identification. Springer-Verlag, New York, 1990.

....their placing. Most of the previous methods are strongly dependent on the amount of textual coverage on the page. Their efficiency usually drops if the text content in page diminishes. This paper proposes a new approach based on texture orientation analysis, earlier used for other purposes [4] and [5], which does not require a text oriented source in order to define the skew error: a few lines of text or pictures that usually have the same orientation as the text are sufficient. Keeping in mind that although a text document from a texture point of view is quite complex, it is possible to ....

....the shrinking factor used with this algorithm is small enough not to cause any loss to the texture look of textual contents in the image. 2. 2 Texture direction analysis The actual texture direction analysis is based on the directional evidence accumulation presented by Chaudhuri [4] and Rao [5]. In this method, the edge image is formed first. After that the local dominant orientations are computed for non overlapping subwindows and used to increment a histogram of orientations. Finally, the maximum peaks are detected from the histogram, in which the highest peak defines directly the ....

Rao A. (1990) A Taxonomy for Texture Description and Identification. Springer, New York.

....image represents an intrinsic property of the fingerprint images and defines invariant coordinates for ridges and furrows in a local neighborhood. By viewing a fingerprint image as an oriented texture, a number of methods have been proposed to estimate the orientation field of fingerprint images [11, 16, 10, 1]. We have developed a least mean square orientation estimation algorithm. Given a normalized image, G, the main steps of the algorithm are as follows: 1. Divide G into blocks of size w Theta w (16 Theta 16) 2. Compute the gradients x (i; j) and y (i; j) at each pixel, i; j) Depending on ....

....local ridge frequency is another intrinsic property of a fingerprint image. Let G be the normalized image and O be the orientation image, then the steps involved in local ridge frequency estimation are as follows: a) b) Figure 6: Comparison of orientation fields by the method proposed in [16] and our method; w = 16 and w Phi = 5. X a x signature Oriented Window Local Ridge Oirentation Block b Figure 7: Oriented window and x signature. 1. Divide G into blocks of size w Theta w (16 Theta 16) 2. For each block centered at pixel (i; j) compute an oriented window of size l Theta ....

A. R. Rao. A Taxonomy for Texture Description and Identification. Springer-Verlag, New York, 1990.

....of Orientation Field A new hierarchical orientation field estimation algorithm has been implemented to estimate the orientation field of an input fingerprint image. It consists of the following two main steps: 1. Estimate the local orientation at each pixel (i; j) using the approach in [11]: 2. Compute the variance of the orientation field in a local neighborhood at each pixel (i; j) If it is above a certain threshold T v , then the local orientation at this pixel is re estimated at a lower resolution level until it is above the threshold value. Experimental results show that even ....

A. Ravishankar Rao, A Taxonomy for Texture Description and Identification, Springer-Verlag, New York, 1990.

No context found.

A.R. Rao, A Taxonomy for Texture Description and Identification, Springer Verlag 1990.

No context found.

A. R. Rao. A taxonomy for texture description and identification. Springer, 1990.

No context found.

A. R. Rao. A Taxonomy for Texture Description and Identification. Springer-Verlag, 1990.

No context found.

A.R. Rao. A Taxonomy for Texture Description and Identification. Springer-Verlag, 1990.

No context found.

A. R. Rao. A taxonomy for texture description and identification. Springer-Verlag, 1990.

No context found.

A. Ravishankar Rao. A Taxonomy for Texture Description and Identification. Springer-Verlag, 1990.

No context found.

A. Ravishankar Rao, A Taxonomy for Texture Description and Identification, Springer-Verlag, New York #1990#.

No context found.

A.Rao "A taxonomy for texture description and identification", Springer-Verlag, New York, 1990

No context found.

A. Ravishankar Rao, A Taxonomy for Texture Description and Identification, Springer-Verlag, New York, 1990.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC