scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

Rapid Texture Identification

Kenneth I. Laws1
23 Dec 1980-Vol. 0238, pp 376-381
TL;DR: In this article, the texture energy approach requires only a few convolutions with small (typically 5x5) integer coefficient masks, followed by a moving-window absolute average operation.
Abstract: A method is presented for classifying each pixel of a textured image, and thus for segmenting the scene. The "texture energy" approach requires only a few convolutions with small (typically 5x5) integer coefficient masks, followed by a moving-window absolute average operation. Normalization by the local mean and standard deviation eliminates the need for histogram equalization. Rotation-invariance can also be achieved by using averages of the texture energy features. The convolution masks are separable, and can be implemented with 1-dimensional (vertical and horizontal) or multipass 3x3 convolutions. Special techniques permit rapid processing on general-purpose digital computers.
Citations
More filters
Journal ArticleDOI
TL;DR: The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) as mentioned in this paper was organized in conjunction with the MICCAI 2012 and 2013 conferences, and twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low and high grade glioma patients.
Abstract: In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients—manually annotated by up to four raters—and to 65 comparable scans generated using tumor image simulation software Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%–85%), illustrating the difficulty of this task We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource

3,699 citations

Book
01 Dec 1993
TL;DR: The geometric, random field, fractal, and signal processing models of texture are presented and major classes of texture processing such as segmentation, classification, and shape from texture are discussed.
Abstract: This chapter reviews and discusses various aspects of texture analysis. The concentration is o the various methods of extracting textural features from images. The geometric, random field, fractal, and signal processing models of texture are presented. The major classes of texture processing pro lems such as segmentation, classification, and shape from texture are discussed. The possible applic tion areas of texture such as automated inspection, document processing, and remote sensing a summarized. A bibliography is provided at the end for further reading.

2,257 citations

Journal ArticleDOI
12 Dec 2017-JAMA
TL;DR: In the setting of a challenge competition, some deep learning algorithms achieved better diagnostic performance than a panel of 11 pathologists participating in a simulation exercise designed to mimic routine pathology workflow; algorithm performance was comparable with an expert pathologist interpreting whole-slide images without time constraints.
Abstract: Importance Application of deep learning algorithms to whole-slide pathology images can potentially improve diagnostic accuracy and efficiency. Objective Assess the performance of automated deep learning algorithms at detecting metastases in hematoxylin and eosin–stained tissue sections of lymph nodes of women with breast cancer and compare it with pathologists’ diagnoses in a diagnostic setting. Design, Setting, and Participants Researcher challenge competition (CAMELYON16) to develop automated solutions for detecting lymph node metastases (November 2015-November 2016). A training data set of whole-slide images from 2 centers in the Netherlands with (n = 110) and without (n = 160) nodal metastases verified by immunohistochemical staining were provided to challenge participants to build algorithms. Algorithm performance was evaluated in an independent test set of 129 whole-slide images (49 with and 80 without metastases). The same test set of corresponding glass slides was also evaluated by a panel of 11 pathologists with time constraint (WTC) from the Netherlands to ascertain likelihood of nodal metastases for each slide in a flexible 2-hour session, simulating routine pathology workflow, and by 1 pathologist without time constraint (WOTC). Exposures Deep learning algorithms submitted as part of a challenge competition or pathologist interpretation. Main Outcomes and Measures The presence of specific metastatic foci and the absence vs presence of lymph node metastasis in a slide or image using receiver operating characteristic curve analysis. The 11 pathologists participating in the simulation exercise rated their diagnostic confidence as definitely normal, probably normal, equivocal, probably tumor, or definitely tumor. Results The area under the receiver operating characteristic curve (AUC) for the algorithms ranged from 0.556 to 0.994. The top-performing algorithm achieved a lesion-level, true-positive fraction comparable with that of the pathologist WOTC (72.4% [95% CI, 64.3%-80.4%]) at a mean of 0.0125 false-positives per normal whole-slide image. For the whole-slide image classification task, the best algorithm (AUC, 0.994 [95% CI, 0.983-0.999]) performed significantly better than the pathologists WTC in a diagnostic simulation (mean AUC, 0.810 [range, 0.738-0.884];P Conclusions and Relevance In the setting of a challenge competition, some deep learning algorithms achieved better diagnostic performance than a panel of 11 pathologists participating in a simulation exercise designed to mimic routine pathology workflow; algorithm performance was comparable with an expert pathologist interpreting whole-slide images without time constraints. Whether this approach has clinical utility will require evaluation in a clinical setting.

2,116 citations

Journal ArticleDOI
TL;DR: Most major filtering approaches to texture feature extraction are reviewed and a ranking of the tested approaches based on extensive experiments is presented, showing the effect of the filtering is highlighted, keeping the local energy function and the classification algorithm identical for most approaches.
Abstract: In this paper, we review most major filtering approaches to texture feature extraction and perform a comparative study. Filtering approaches included are Laws masks (1980), ring/wedge filters, dyadic Gabor filter banks, wavelet transforms, wavelet packets and wavelet frames, quadrature mirror filters, discrete cosine transform, eigenfilters, optimized Gabor filters, linear predictors, and optimized finite impulse response filters. The features are computed as the local energy of the filter responses. The effect of the filtering is highlighted, keeping the local energy function and the classification algorithm identical for most approaches. For reference, comparisons with two classical nonfiltering approaches, co-occurrence (statistical) and autoregressive (model based) features, are given. We present a ranking of the tested approaches based on extensive experiments.

1,567 citations


Cites background from "Rapid Texture Identification"

  • ...Examples of classification schemes reported in the literature are Bayes classifiers assuming multivariate Gaussian feature distributions [11], [15], [16], [23], Fisher transformation [4], [26], nearest neighbor classifiers [7], [10], [22], classification trees [26], feed-forward neural networks…...

    [...]

  • ...Classification is the task of assigning class labels to these feature vectors....

    [...]

  • ...F ——————————...

    [...]

  • ...Jain and Farrokhnia [20] suggested a bank of Gabor filters, i.e., Gaussian shaped band-pass filters, with dyadic coverage of the radial spatial frequency range and multiple orientations....

    [...]

Journal ArticleDOI
TL;DR: A modified box-counting approach is proposed to estimate the FD, in combination with feature smoothing in order to reduce spurious regions and to segment a scene into the desired number of classes, an unsupervised K-means like clustering approach is used.
Abstract: This paper deals with the problem of recognizing and segmenting textures in images. For this purpose the authors employ a technique based on the fractal dimension (FD) and the multi-fractal concept. Six FD features are based on the original image, the above average/high gray level image, the below average/low gray level image, the horizontally smoothed image, the vertically smoothed image, and the multi-fractal dimension of order two. A modified box-counting approach is proposed to estimate the FD, in combination with feature smoothing in order to reduce spurious regions. To segment a scene into the desired number of classes, an unsupervised K-means like clustering approach is used. Mosaics of various natural textures from the Brodatz album as well as microphotographs of thin sections of natural rocks are considered, and the segmentation results to show the efficiency of the technique. Supervised techniques such as minimum-distance and k-nearest neighbor classification are also considered. The results are compared with other techniques. >

650 citations

References
More filters
Journal ArticleDOI
01 Nov 1973
TL;DR: These results indicate that the easily computable textural features based on gray-tone spatial dependancies probably have a general applicability for a wide variety of image-classification applications.
Abstract: Texture is one of the important characteristics used in identifying objects or regions of interest in an image, whether the image be a photomicrograph, an aerial photograph, or a satellite image. This paper describes some easily computable textural features based on gray-tone spatial dependancies, and illustrates their application in category-identification tasks of three different kinds of image data: photomicrographs of five kinds of sandstones, 1:20 000 panchromatic aerial photographs of eight land-use categories, and Earth Resources Technology Satellite (ERTS) multispecial imagery containing seven land-use categories. We use two kinds of decision rules: one for which the decision regions are convex polyhedra (a piecewise linear decision rule), and one for which the decision regions are rectangular parallelpipeds (a min-max decision rule). In each experiment the data set was divided into two parts, a training set and a test set. Test set identification accuracy is 89 percent for the photomicrographs, 82 percent for the aerial photographic imagery, and 83 percent for the satellite imagery. These results indicate that the easily computable textural features probably have a general applicability for a wide variety of image-classification applications.

20,442 citations


Additional excerpts

  • ...Svh(r,c) = (14/15) Svh(r,c) + lEvh(r,c) (5)...

    [...]

ReportDOI
01 Jan 1980
TL;DR: In this article, texture energy is measured by filtering with small masks, typically 5x5, then with a moving-window average of the absolute image values, leading to a simple class of texture energy transforms, which perform better than any of the preceding methods.
Abstract: : The problem of image texture analysis is introduced, and existing approaches are surveyed. An empirical evaluation method is applied to two texture measurement systems, co-occurrence statistics and augmented correlation statistics. A spatial-statistical class of texture measures is then defined and evaluated. It leads to a simple class of texture energy transforms, which perform better than any of the preceding methods. These transforms are very fast, and can be made invariant to changes in luminance, contrast, and rotation without histogram equalization or other preprocessing. Texture energy is measured by filtering with small masks, typically 5x5, then with a moving-window average of the absolute image values. This method, similar to human visual processing, is appropriate for textures with short coherence length or correlation distance. The filter masks are integer-valued and separable, and can be implemented with one-dimensional or 3x3 convolutions. The averaging operation is also very fast, with computing time independent of window size. Texture energy planes may be linearly combined to form a smaller number of discriminant planes. These principal component planes seem to represent natural texture dimensions, and to be more reliable texture measures than the texture energy planes. Texture segmentation or classification may be accomplished using either texture energy or principal component planes as input. This study classified 15x15 blocks of eight natural textures. Accuracies of 72% were achieved with co- occurrence statistics, 65% with augmented correlation statistics, and 94% with texture energy statistics.

869 citations


Additional excerpts

  • ...c+7 r+7 Evh(r,c) = j=c-7 i=r-7 Fvh(i,j) (1)...

    [...]