Mikeš, Stanislav · Digital Repository (National Repository of Grey Literature) · 2003
An efficient and robust type of unsupervised multispectral texture segmentation method is presented. The algorithm starts with spectral factorization of an input multispectral texture image using the Karhunen-Loeve expansion. Monospectral factors of single texture patches are assumed to be modelled using a Gaussian Markov random field model. The texture segmentation is done by K-means algorithm in the Markov model parameter space evaluated for each pixel centered image window.
Cited by 0
Michal Haindl, Stanislav Mikeš · InTech eBooks · 2008
We discussed three efficient and robust methods for unsupervised texture segmentation with unknown number of classes based on the underlying Markovian and GM texture models and their modifications for medical mammographics and remote sensing applications, respectively. Although these algorithm use the random field type models they are fast because they use efficient recursive or pseudo-likelihood parameter estimation of the underlying texture models and therefore they are much faster than the usual Markov
Cited by 3Open access
Reyer Zwiggelaar, Erika Denton · Lecture notes in computer science · 2006
Cited by 13
Gianfranco Doretto, Daniel Cremers, Paolo Favaro et al. · Open Access CRIS of the University of Bern · 2017
We address the problem of segmenting a sequence of im-ages of natural scenes into disjoint regions that are charac-terized by constant spatio-temporal statistics. We model the spatio-temporal dynamics in each region by Gauss-Markov models, and infer the model parameters as well as the boundary of the regions in a variational optimization frame-work. Numerical results demonstrate that – in contrast to purely texture-based segmentation schemes – our method is effective in segmenting regions that differ in their dynamics even when spatial statistics are identical. 1.
Cited by 38Open access
Lech Szumilas, Allan Hanbury · 2006
A novel approach to automatic, hierarchical texture detection in single images as a step towards image understanding is presented. First, the proposed method searches for alternating color patterns through hierarchical clustering of color pairs from adjacent, symmetrical image segments to localize salient regions in terms of color and texture. Second, the salient regions are fed as seeds to an image segmentation method based on min-cut/max-flow in the graph to localize the texture boundaries more accurately. The final result is a hierarchy of potential textured regions in the image useful for further object/texture recognition step. This work gives a proof of concept that the stable salient texture regions supported by a semi-automatic segmentation algorithm may provide fully automatic image segmentation into uniform color and/or texture regions. The results are presented on some images of natural scenes from the Berkeley database.
Cited by 3
Sarah Harrison, David R. T. Keeble · Ophthalmic and Physiological Optics · 2002
Cited by 3
Sarah Harrison, David R. T. Keeble · Vision Research · 2008
Cited by 13
Ahsan Ahmad Ursani, A. W. Umrani, Fahim Aziz Umrani · SHILAP Revista de lepidopterología · 2010
Texture is an important image-content that has been utilized for different machine intelligent tasks, like those in machine vision and remote sensing, which identify objects of interest by segmenting the image texture. This paper aims at comparing texture features based on Discrete Fourier Transform (DFT) with ones based on Gabor wavelets for unsupervised image segmentation. The comparison is realized theoretically, analytically, as well as empirically. Images of natural scenes from a standard image database have been taken as test images. Analytical comparison shows that the DFT-based features are computationally less expensive than those based on Gabor wavelets. Empirical results show that the performance of the texture features based on DFT is comparable to those based on Gabor wavelets.
Cited by 0Open access
Pasáček, Václav · Brno University of Technology Digital Library (Brno University of Technology) · 2012
Image segmentation is an important step in image processing. A traditional way how to segment an image is a texture-based segmentation that uses texture features to describe image texture. In this work, Local Binary Patterns (LBP) are used for image texture representation. Texture feature is a histogram of occurences of LBP codes in a small image window. The work also aims to comparison of results of various modifications of Local Binary Patterns and their usability in the image segmentation which is done by unsupervised clustering of texture features. The Fuzzy C-Means algorithm is finally used for the clustering in this work.
Cited by 0
M. Wilscy, Remya K. Sasi · 2010
Hierarchical Texture Segmentation using Wavelet packet decomposition performs an unsupervised classification of Texture features, extracted using wavelet packet decomposition to generate a segmented image. Recursive decomposition of both the approximation and the detail coefficients derived from the original signal provides a wider spectrum for richer feature extraction. This Texture Segmentation works well for Texture images with only two Textures. This approach can be used for shape extraction, detecting industrial defects, and for text extraction.
Cited by 3
L.M. Linnett · 1991
Cited by 8
Reyer Zwiggelaara, Lilian Blota, David Rabab et al. · 2015
Abstract. We have investigated a combination of statistical modelling and expectation maximisation for a texture based approach to the segmentation of mammographic images. Texture modelling is based on the implicit incorporation of spatial information through the introduction of a set-permutation-occurrence matrix. Statistical modelling is used for dimensionality reduction, data generalisation and noise removal purposes. Expectation maximisation modelling of the resulting feature vector provides the basis for image segmentation. The developed segmentation results are used for automatic mammographic risk assessment. 1
Cited by 0
Oriol Pujol, et al. · 2007
... In this paper we propose a new deformable model defined in a statistic framework to segment objects of natural scenes. We perform a supervised learning of local appearance of the textured objects and construct a feature space using a set of co-occurrence matrix measures. Linear Discriminant Analysis allows us to obtain an optimal reduced feature space where a mixture model is applied to construct a likelihood map. Instead of using a heuristic potential field, our active model is deformed on a regularized version of the likelihood map in order to segment objects characterized by the same texture pattern. Different tests on synthetic images, natural scene and medical images show the advantages of our statistic deformable model.
Cited by 0
Mihran Tüceryan · 2003
Texture segmentation is one of the early steps towards identifying surfaces and objects in an image. In the paper a moment based texture segmentation algorithm is presented. The moments in small windows of the image are used as texture features which are then used to segment the textures. The algorithm has successfully segmented binary images containing textures with identical second-order statistics as well as a number of natural gray level texture images.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>
Cited by 32
Ivan Ustyuzhaninov, Claudio Michaelis, Wieland Brendel et al. · arXiv (Cornell University) · 2018
We introduce one-shot texture segmentation: the task of segmenting an input image containing multiple textures given a patch of a reference texture. This task is designed to turn the problem of texture-based perceptual grouping into an objective benchmark. We show that it is straight-forward to generate large synthetic data sets for this task from a relatively small number of natural textures. In particular, this task can be cast as a self-supervised problem thereby alleviating the need for massive amounts of manually annotated data necessary for traditional segmentation tasks. In this paper we introduce and study two concrete data sets: a dense collage of textures (CollTex) and a cluttered texturized Omniglot data set. We show that a baseline model trained on these synthesized data is able to generalize to natural images and videos without further fine-tuning, suggesting that the learned image representations are useful for higher-level vision tasks.
Cited by 11Open access
Greg Appelbaum, Alex R. Wade, Mark W. Pettet et al. · Journal of Vision · 2005
The segmentation of visual images into coherent objects and their surrounds is a fundamental process of the early visual system. To assess the role of local boundary cues in figure-ground segmentation, we constructed displays in which figure and background regions were separately ‘tagged’ with periodic modulations of their texture orientations or phase (relative alignment). Tagging allowed us to monitor figure and background responses separately and to study non-linear interactions between them as a function of segmentation cue and retinotopic location. Steady-state visually evoked potentials (SSVEPs) were recorded over 128 channels as subjects passively viewed the stimuli. Distributed source modeling was performed based on individual head models and cortical current density distributions were plotted in relationship to visual areas defined by fMRI retinotopic maps. SSVEP responses at harmonics of the figure tag-frequency were maximal over lateral regions of the occipital cortex, while harmonic responses of the background tag-frequency were restricted to the midline occipital pole. These responses were typically cue and retinotopically invariant. Responses at frequencies equal to low-order sums and differences of the tag frequencies, resulting from non-linear interactions occurring between the figure and background regions, were both cue and retinotopically dependent.
Cited by 0Open access
Jiang Wu, Chunxiao Liu · Journal of Image and Graphics · 2021
目的超像素分割是计算机视觉领域常用的一项预处理技术，目标是将相邻像素聚集成为具有一定语义的子区域，能够大幅度降低后续处理的计算复杂度，但是对包含强梯度纹理的图像分割效果不佳，为此提出一种具有纹理感知能力的超像素分割方法。方法提出一种能够区分强梯度噪声和纹理像素的颜色距离，其中利用带方向的1/4圆形窗口均值滤波后的颜色信息，提升包含强梯度噪声和纹理图像的超像素分割性能。利用区间梯度幅值与Sobel梯度幅值相乘得到混合梯度幅值，具有纹理抑制、结构保持以及边缘线条细的优点，能够提升超像素的贴合边缘性能，增强超像素形状规则程度。最后，利用混合梯度的幅值计算具有结构回避能力的综合聚类距离，进一步防止超像素跨越物体的边界，增强超像素的贴边性能。结果在BSDS500(Berkeley segmentation dataset 500)图像数据集和强纹理马赛克图像等不同类型图像上的测试结果显示，与目前主流的超像素分割方法相比，本文算法在UE (undersegmentation error)、ASA (achievable segmentation accuracy)和CM (compactness measure)等性能指标上分别提高了1.5%、0.2%和4.3%。从视觉效果上看，能够在排除纹理干扰的情况下生成结构边缘贴合程度更好的形状规则超像素。结论本文算法在包含强梯度纹理图像上的超像素分割性能优于对比方法，在目标识别、目标追踪和显著性检测等易受强梯度干扰的技术领域具有较大应用潜力。
Cited by 1
Neo, Siong Huat. · DR-NTU (Nanyang Technological University) · 1999
The focus of this thesis is on the development of texture segmentation algorithm. The texture features are extracted using multi-channel Gabor filters. The ISODATA clustering algorithm is implemented for texture features discrimination.
Cited by 4
Oscar Nestares And, Oscar Nestares, Javier Portilla · 2007
echniques for spatial varying signals, like images containing several textured regions. The main steps of the proposed method for texture segmentation are displayed in the diagram of Figure 1. The texture descriptors at each location are the weights of the AR filter, reflecting the correlation present at different orientations (i.e., the directionality) and the absolute value of the prediction error, that reflects the regularity of the textures. We have applied a very efficient adaptive filtering using a small mask (four coefficients), checking three algorithms for the weight adaptation. The results, showed in Figure 2, are highly satisfactory with test images containing textures from the Brodatz&apos;s album, with mean percentages of correctly segmented pixels higher than 94%. These results permit the following ordering in increasing performance and also in computational cost: 1) The LLMS algorithm provided reasonable results at very low computational cost; 2) The one-dimen
Cited by 0
B.M. Waller, Mark Nixon, John D. Carter · ePrints Soton (University of Southampton) · 2011
A new approach to texture segmentation is presented which uses Local Binary Pattern data to provide evidence from which pixels can be classified into texture classes. The proposed algorithm, which we contend to be the first use of evidence gathering in the field of texture classification, uses Generalised Hough Transform style R-tables as unique descriptors for each texture class and an accumulator is used to store votes for each texture class. Tests on the Brodatz database and Berkeley Segmentation Dataset have shown that our algorithm provides excellent results; an average of 86.9% was achieved over 50 tests on 27 Brodatz textures compared with 80.3% achieved by segmentation by histogram comparison centred on each pixel. In addition, our results provide noticeably smoother texture boundaries and reduced noise within texture regions. The concept is also a "higher order" texture descriptor, whereby the arrangement of texture elements is used for classification as well as the frequency of occurrence that is featured in standard texture operators. This results in a unique descriptor for each texture class based on the structure of texture elements within the image, which leads to a homogeneous segmentation, in boundary and area, of texture by this new technique.
Cited by 1Open access