The statistics of gray level differences have been successfully used in a number of texture analysis studies. In this paper we propose to use signed gray level differences and their multidimensional distributions for texture description. The present approach has important advantages compared to earlier related approaches based on gray level cooccurrence matrices or histograms of absolute gray level differences. Experiments with a difficult texture classification problem show that our approach provides a very good and robust classification performance in comparison to the mainstream paradigms such as cooccurrence matrices, Gaussian Markov Random Fields, or Gabor filtering.

We consider the problem of zone classification in document image processing. Document blocks are labelled as text or non-text using texture features derived from a feature based interaction map (FBIM), a recently introduced general tool for texture analysis [3, 4]. The zone classification procedure proposed is tested on the comprehensive document image database UW-I created at the University of Washington in Seattle. Different classification procedures are considered. The performance ranges from 96 % to 98 % using 6 FBIM texture features only. 1 1. Introduction Document image understanding involves determining the geometric page layout, labeling blocks as text or nontext, determining the read order for text blocks, recognizing the text of text blocks through an OCR system, determining the logical page layout, and formatting the data and information of the document in a suitable way for use by a word processing system or by an information retrieval system [5]. The zone classification ...

We discuss a novel method for recovering fundamental, perceptually motivated structural features of a texture pattern: anisotropy, symmetry, and regularity. The method is based on extended spatial grey-level difference statistics which describe pairwise pixel interactions and yield an interaction map used to assess the overall two-dimensional structure of interactions and extract the significant short- and long-range interactions (intersample spacings). The new approach extends, in digital images, the notion of greylevel difference to arbitrary spacing vectors (i.e. any angle at any displacement). This provides the necessary background for precise anisotropy (or directionality) and symmetry analysis. Experimental results are shown with a set of Brodatz images that range from highly regular to patterns with weak regularity or anisotropy. A few especially interesting examples of recovering hardly visible structural features are given. Finally, the approach is applied to rotation-invariant texture classification.

: Human texture perception relies on a few basic high-level features including directionality and symmetry. Today, research on oriented patterns finds its applications in various areas of applied machine vision. In this study, we present and investigate a new method for assessing pattern anisotropy via texture symmetry. Pattern orientation is viewed as the direction of persistent statistical texture symmetry. The proposed method uses the spatial dependence of an extended spatial gray-level difference feature to yield an interaction symmetry map which reflects the symmetry of both short- and long-range pixel interactions. Pattern orientation can then be assessed via the directions of global symmetry - the characteristic axes of the pattern. Experimental results are shown which support our claim that texture symmetry is deeply related to the perceived orientation. The results are compared to the orientations obtained in a recent study that uses the traditional filtering framework. The pr...

We have recently introduced a new tool for texture analysis called feature based interaction map (FBIM). The FBIM approach can be efficiently used to assess fundamental structural properties of textures such as anisotropy, symmetry, orientation and regularity [4]. It has been demonstrated [5] that the FBIM is suitable for rotation-invariant texture classification of patterns with regular, weak regular, or linear structure. In this paper, we show how the interaction map can be applied as a structural filter for segmentation, detection of textured objects and texture defects, analysis of oriented structures and shape-from-texture. The power of the FBIM filter is in its unique capability to grasp the structure of pixel interactions typical for a given texture pattern. To efficiently use this capability, we propose a fast running implementation of the FBIM algorithm and present pilot experimental results demonstrating the potential of the FBIM approach in diverse tasks and applications. 1...

Pairwise pixel interactions have proved to be a powerful tool in feature based (8, 11) and model based (13, 14) texture analysis. Various successful applications of the feature based interaction map (FBIM) approach have already been presented. (8, 9, 11, 12) Different aspects and components of the approach have been discussed, but no self-contained description of the FBIM method has been published yet. This paper provides a comprehensive up-to-date survey of the approach, including all major algorithms and a series of systematic experimental studies that demonstrate the capabilities of the approach. Texture analysis, Feature based interaction map, Anisotropy, Regularity, Symmetry, Pattern orientation To appear in: Pattern Recognition, Special Issue on Color and Texture, 1 Introduction Motivated by the discovery of the high level texture features responsible for perceptual grouping of textures (21) and the development of the Markov-Gibbs texture model with pairwise pixel i...

Introduction Motivated by the discovery of the high level texture features responsible for perceptual grouping of textures [11] and the development of the Markov-Gibbs texture model with pairwise pixel interactions [9], we have recently proposed the method of feature based interaction maps (FBIM) and applied this new tool to the problem of pattern orientation [4] and rotation-invariant texture classification [7]. Experimental results have demonstrated that the FBIM approach can be used to recover the basic structural properties and orientation of a wide range of patterns, including weak structures. In [11], the fundamental, perceptually motivated high level texture features were identified as directionality (anisotropy) versus nondirectionality, periodicity versus irregularity and, probably, structural complexity. These features reflect the intrinsic symmetry properties of the process, natural or artificial, that generates the texture pattern. Figure 1 illustra

In this paper, we consider the problem of automatically detecting a facial symmetry axis in what we will call a standard human face image (acquired when the subject is looking directly into the camera, in front of a clean gray background under controlled illumination). Images of this kind are encountered in face recognition scenarios; this detection should facilitate more sophisticated facial image processing. The proposed method is based on GLDH (gray level difference histogram) analysis and consists of three components: (1) the face region detection stage crops an approximate face region out of the background, (2) symmetry detection discovers a vertical axis to optimally bisect the region of interest, assuming it is bilaterally symmetric, and (3) orientation adjustment aligns the angle of the symmetry axis with the orientation of the face. An implementation of the method is described, and results are presented. This detector’s robust performance is evidenced by its success finding symmetry axes in more than 7,500 images collected from 600 distinct subjects. One of our method’s most noteworthy contributions is that, according to our experimental results, many of the automatically detected axes are more accurate than the reference axes. Our automated detector is a powerful tool because it is not as susceptible to human error as its manual counterpart and, as the first application of its kind, it could potentially serve as a new biometric. 1.

We have recently introduced a new tool for texture analysis called feature based interaction map (FBIM). The FBIM approach can be efficiently used to assess fundamental structural properties of textures such as anisotropy, symmetry, orientation and regularity [4]. It has been demonstrated [5] that the FBIM is suitable for rotation-invariant texture classification of patterns with regular, weak regular, or linear structure. In this paper, we show how the interaction map can be applied as a structural filter for segmentation, detection of textured objects and texture defects, analysis of oriented structures and shape-from-texture. The power of the FBIM filter is in its unique capability to grasp the structure of pixel interactions typical for a given texture pattern. To efficiently use this capability, we propose a fast running implementation of the FBIM algorithm and present pilot experimental results demonstrating the potential of the FBIM approach in diverse tasks and applications. 1 1.

Abstract—A method is proposed for three-dimensional (3-D) texture analysis of magnetic resonance imaging brain datasets. It is based on extended, multisort co-occurrence matrices that employ intensity, gradient and anisotropy image features in a uniform way. Basic properties of matrices as well as their sensitivity and dependence on spatial image scaling are evaluated. The ability of the suggested 3-D texture descriptors is demonstrated on nontrivial classification tasks for pathologic findings in brain datasets. Index Terms—3-D texture, co-occurrence, MRI, neurodegenerative diseases.