Search

Local Binary Pattern (LBP) is a feature extraction operator with both high texture discrimination ability and low computational complexity. Many LBP variants have been proposed to improve the performance of texture classification or overcome the drawbacks of LBP. There are three shortcomings in some LBP variants: discarding the magnitude component between local differences, adopting fixed weights in the encoding process and discarding the absolute information of the pixel gray level. Based on the three points, this paper proposes an improved LBP with two operators, local binary pattern operator based on magnitude ranking and global threshold segmentation operator, to further improve the performance. This improved LBP can achieve excellent texture classification accuracy across six common datasets, with an average of 1% lower than the best LBP variants. Meanwhile, the computational complexity of the proposed improved LBP is several times lower than that of the best LBP variants. [ABSTRACT FROM AUTHOR]

Local binary pattern is one of the most known descriptors, which is used for texture classification. Although completed local binary pattern is seemingly the most precise variant of this type of descriptor and provides high classification accuracy by joining three histograms of features. Merging these histograms increases the features number significantly. To reduce the size of features, in this paper, some mapping methods are proposed for feature reduction and mapping of these features into a histogram. All of the proposed mapping methods are rotation and illumination invariant. Furthermore, a constraint feature selection method is proposed that selects discriminative features. Applying the introduced methods to the known benchmarks like Outex (TC3, TC10, TC13, TC12(t) and TC12(h)), UIUC, CUReT and Defect Fabric datasets indicates that even by adopting lower number of features, the classification rate is enhanced from 1% to 9% while the features number are decreased around 10% to 99%. Comparison results on the same datasets imply the superiority of the proposed schemes to the conventional methods. [ABSTRACT FROM AUTHOR]

The analysis of textures is an important task in image processing and computer vision because it provides significant data for image retrieval, synthesis, segmentation, and classification. Automatic texture recognition is difficult, however, and necessitates advanced computational techniques due to the complexity and diversity of natural textures. This paper presents a method for classifying textures using graphs; specifically, natural and horizontal visibility graphs. The related image natural visibility graph (INVG) and image horizontal visibility graph (IHVG) are used to obtain features for classifying textures. These features are the clustering coefficient and the degree distribution. The suggested outcomes show that the aforementioned technique outperforms traditional ones and even comes close to matching the performance of convolutional neural networks (CNNs). Classifiers such as the support vector machine (SVM), K-nearest neighbor (KNN), decision tree (DT), and random forest (RF) are utilized for the categorization. The suggested method is tested on well-known image datasets like the Brodatz texture and the Salzburg texture image (STex) datasets. The results are positive, showing the potential of graph methods for texture classification. [ABSTRACT FROM AUTHOR]

Our world consists not only of objects and scenes but also of materials of various kinds. Being able to recognize the materials that surround us (e.g., plastic, glass, concrete) is important for humans as well as for computer vision systems. Unfortunately, materials have received little attention in the visual recognition literature, and very few computer vision systems have been designed specifically to recognize materials. In this paper, we present a system for recognizing material categories from single images. We propose a set of low and mid-level image features that are based on studies of human material recognition, and we combine these features using an SVM classifier. Our system outperforms a state-of-the-art system (Varma and Zisserman, TPAMI 31(11):2032-2047, ) on a challenging database of real-world material categories (Sharan et al., J Vis 9(8):784-784a, ). When the performance of our system is compared directly to that of human observers, humans outperform our system quite easily. However, when we account for the local nature of our image features and the surface properties they measure (e.g., color, texture, local shape), our system rivals human performance. We suggest that future progress in material recognition will come from: (1) a deeper understanding of the role of non-local surface properties (e.g., extended highlights, object identity); and (2) efforts to model such non-local surface properties in images. [ABSTRACT FROM AUTHOR]

High-efficiency video coding (HEVC/H.265) is one of the most widely used video coding standards. HEVC introduces a quad-tree coding unit (CU) partition structure to improve video compression efficiency. The determination of the optimal CU partition is achieved through the brute-force search rate-distortion optimization method, which may result in high encoding complexity and hardware implementation challenges. To address this problem, this paper proposes a method that combines convolutional neural networks (CNN) with joint texture recognition to reduce encoding complexity. First, a classification decision method based on the global and local texture features of the CU is proposed, efficiently dividing the CU into smooth and complex texture regions. Second, for the CUs in smooth texture regions, the partition is determined by terminating early. For the CUs in complex texture regions, a proposed CNN is used for predictive partitioning, thus avoiding the traditional recursive approach. Finally, combined with texture classification, the proposed CNN achieves a good balance between the coding complexity and the coding performance. The experimental results demonstrate that the proposed algorithm reduces computational complexity by 61.23%, while only increasing BD-BR by 1.86% and decreasing BD-PSNR by just 0.09 dB. [ABSTRACT FROM AUTHOR]

The main focus of this paper is to improve the performance of the texture model Optimized Local Ternary Patterns (OLTP), which is known as one of the successful variants of the texture model Local Binary Patterns (LBP), a well known method for texture analysis and its applications. Generally preprocessing is used in digital image processing for reducing the unwanted noise and disturbances in such a way that it improves the quality of the image. Preprocessing not only removes the distortions but also enhances the features of the image for further processing. To achieve better recognition accuracy, the texture model OLTP was combined with a preprocessing method that uses nonlinear diffusion method as a preprocessing tool in this paper, with the hope that this idea will surely improve the local feature description and texture classification process. This nonlinear diffusion method uses two newly developed edge stopping functions for preprocessing. This proposed method is tested with two standard texture datasets namely Brodatz Dataset and Usptex dataset. The results show that the use of the preprocessing step really improved the texture classification accuracy. [ABSTRACT FROM AUTHOR]

In this paper, an extended complete LBP (ELBP) for texture classification is proposed, in which the local feature vectors are composed of the ratio of the central pixel and its neighborhood pixels to a specific threshold. ECLBP_C represents the gray level of the image, which is obtained by comparing the center pixel with the global threshold. ECLBP_S and ECLBP_M represent the symbol component and the magnitude component of the 3-neighbor region of the center pixel respectively, which are obtained by calculating two binary codes using the original LBP algorithm for the 3-neighbor region of the center pixel. In order to make the proposed algorithm scalable, in addition to the 3-neighbor pixels of the central pixel, the proposed algorithm use the center pixel as the center, r as the radius in the circle with ɑ as the filter's radius to generate extended binary coding, such as ECLBP_ES_r,α ECLBP_EM_ r,α. In order to describe the local region feature vector in detail, specified ECLBP_ES_r,α and ECLBP_EM_r,α can be obtained by defining the number of extensions according to actual needs, and then established and concatenated all ECLBP gray histograms for statistics. In the experimental part, we analyze the performance of the proposed algorithm in detail, and prove that the algorithm has good scalability and robustness. The experimental results show that the classification accuracy of the proposed algorithm is up to 99% after 3 expansions in Table 2. The source codes of the proposed algorithm can be downloaded from https://github.com/zenqiang/ECLBP.git. [ABSTRACT FROM AUTHOR]

Land cover (LC) classification remains a challenging task due to the diversity of terrain and topography, limited prior knowledge, and complex data sets. These variations can lead to differences in illumination and shading, which can affect the appearance of objects in the image. The local binary pattern (LBP) model is an effective technique for capturing local texture information in an image, which can help to overcome the effects of topography diversity by analysing patterns in different image regions, even if the illumination or shading conditions are different. However, LBP alone is inadequate for characterizing high-resolution remotely sensed images with complex semantic content as it only utilizes sign information in the local region. In this paper, a new texture characterization descriptor, known as completed homogeneous LBP (CHLBP), is proposed as an improved version of homogeneous LBP (HLBP) for LC classification of remotely sensed images. The CHLBP method mainly involves the following steps: first, sign and magnitude information from the HLBP descriptor is extracted, providing an effective alternative to the HLBP complementary contrast measure. Second, for each sign and magnitude function, a new splitting factor δ is used to obtain a depth relationship between the centre and its neighbouring pixels and to enhance noise robustness. Finally, the centre pixels representing the image grey level are also considered to contain discriminative information. The performance of our descriptor is evaluated using four challenging texture databases: Outex (TC10, TC12), Geofen Image Dataset (GID), and large-scale aerial (AID). Extensive experiments were performed using four classifiers: Support Vector Machine (SVM), K-Nearest Neighbours (KNN), Random Forest (RF), and Multi-Layer Perceptron (MLP), demonstrating the effectiveness and robustness of our descriptor against noise and free noise conditions in public remote sensing datasets. [ABSTRACT FROM AUTHOR]

In this paper, a new formula is proposed that uses local entropy for texture feature extraction. This new method is similar to entropy; however, it calculates the local entropy of each local patch of textures. Entropy (ENT) is an attribute that measures the randomness of gray-level distribution of image. Entropy extracts dissimilarity of each local patch. In this paper, local entropy is compared to Local Binary Pattern (LBP) and local variance (VAR). All of these descriptors are rotation invariant and are used for extracting the features from each local neighborhood of textures. In spite of low accuracy of VAR and LBP the performance of ENT does not decrease significantly for noisy textures. In other words, ENT is more robust to noise than VAR and LBP. Implementations on Outex, UIUC, CUReT and MeasTex datasets show that entropy is more accurate than variance and LBP. Similar to VAR and LBP, ENT can be combined with other descriptors to improve the performance of classification. For almost all datasets that are used in implementation part, LBP/ENT is more accurate than LBP/VAR for normal and noisy textures. Also the ENT accuracy outperforms the accuracy of VAR and LBP and most of the advanced noise robust LBP versions for low Signal to Noise Ratio (SNR) values (SNR < 10). ENT feature is a continuous value so it is necessary to quantize to discrete value for histogram. The quantization and train step of ENT is the same as VAR. [ABSTRACT FROM AUTHOR]

Texture classification of images is an important problem in texture analysis, with the goal to assign an unknown sample image to one of a set of known texture classes. Many procedures have been proposed. However, traditional texture classification methods are often complicated and time-consuming. In addition, the whole original data set has to be known in advance. This paper presents a novel approach to texture classification that employs cross-covariance matrices in the compressive measurement domain. Since there is a linear relation between the cross-covariance matrix in the measurement domain and that in the transform domain, the characteristics of the original signals can be reflected through the sampling measurements. By abstracting the measurement-domain cross-covariance matrices of each image class, we build a texton dictionary composed of the cross-covariance vectors and then train each image class according to the texton dictionary. Through comparison with all the training results, test images are classified as the matching one. Experiments show that the texture classification method proposed in this paper has better accuracy, with only 14-50 % processing time than the existing methods. [ABSTRACT FROM AUTHOR]

Whether the products on the shelf are neatly displayed is one of the important factors affecting consumers' desire to buy, which is more obvious in the toy industry. In order to detect the neatness of the toy commodities on the shelf, we propose a method based on texture recognition to measure the neatness of the commodities. Since the whole of neat toy commodities will show regular texture features, in the field of computer vision, how to extract effective texture features from image parameters has always been the focus and difficulty of texture analysis. In this paper, we use a multiquadratic kernel modeling learnable local histogram layer to extract effective texture features and use the convolutional block attention module to filter out features that are more conducive to classification, proposed a new network named histogram residual attention network (HRANet). In the test performance stage, we use the proposed HRANet to test on DTD, MINC-2500, GTOS-mobile, and KTH-T2b datasets, and the accuracy is significantly higher than the original HistRes. Finally, we use the features obtained by HRANet to represent the features of product display and use the cosine distance to measure the similarity of product regions, which effectively quantifies the neatness of product display. [ABSTRACT FROM AUTHOR]

Texture classification under varying illumination conditions is one of the most important challenges. This paper presents a new texture classification approach by taking the combinations of robust illumination normalization techniques applied on gray level run length matrix (GLRLM) for texture features extraction. The purpose of selecting the GRLM, as texture descriptor is that, it extracts information of an image from its gray level runs. A set of consecutive, collinear picture points having the same gray level values is considered as a gray level run. The textured materials usually go through a deep change in their images with variations in illumination and camera pose. For instance, keeping all the parameters fixed but just changing the scale and rotation can result in a completely new texture. Hence, change in gray level values also occurred. Dealing with these variations successfully by utilizing GRLM descriptor for texture classification is the main purpose of this paper. In the suggested approach, 2D wavelet, Tan and Triggs (TT) normalization methods are employed to compensate illumination variations. Experimental results on the Brodatz, VisTex, STex and ALOT databases show that the suggested approach improves the performance significantly as compared to the classical GLRLM descriptor. [ABSTRACT FROM AUTHOR]

Local binary pattern (LBP) is an efficient texture descriptor with increasing applications in machine vision. Notwithstanding the great ability of LBP in revealing texture features of natural images, this descriptor is sensitive to noise, and its accuracy is reduced when applied to noisy images. The two important noise sensitive components in computing LBP, which affect the generated binary patterns, are the central pixel value and the neighboring pixels values, which are used for thresholding. This paper proposes a noise robust texture descriptor that applies a novel mechanism for potential noisy central pixel detection and correction. Moreover, the proposed descriptor uses a new sampling method that corrects potential noisy neighboring pixels by replacing them with the median of their various radii neighboring pixels, which are located in a new arc-shaped structure. Since the proposed median arc center corrected binary pattern (MACCBP) uses pixels related to different radii patterns for code generation, both macro and micro structures participate in thresholding. Hence, performance is increased in both noisy and noiseless environments. Furthermore, the MACCBP applies the idea of completed LBP and extracts magnitude and center information in conjunction with the sign to achieve more noise robustness and classification accuracy. The proposed descriptor is extensively examined in noisy and noise-free experiments using Outex, UIUC, UMD and CUReT datasets. The experimental results show that MACCBP achieves high classification accuracy in the experiments with original images of the datasets, and when additive salt and pepper noise, Gaussian white noise and Gaussian blur are applied to the test images. The MACCBP is compared with its well-known state-of-the-art counterparts in terms of classification accuracy in the experiments. The results obtained from numerous and extensive tests demonstrate that the proposed descriptor is notably superior to its competitors in noisy and noiseless experiments. [ABSTRACT FROM AUTHOR]

Color texture classification aims to recognize patterns by the analysis of their colors and their textures. This process requires using descriptors to represent and discriminate the different texture classes. In most traditional approaches, these descriptors are used with a predefined setting of their parameters and computed from images coded in a chosen color space. The prior choice of a color space, a descriptor and its setting suited to a given application is a crucial but difficult problem that strongly impacts the classification results. To overcome this problem, this paper proposes a color texture representation that simultaneously takes into account the properties of several settings from different descriptors computed from images coded in multiple color spaces. Since the number of color texture features generated from this representation is high, a dimensionality reduction scheme by clustering-based sequential feature selection is applied to provide a compact hybrid multi-color space (CHMCS) descriptor. The experimental results carried out on five benchmark color texture databases with five color spaces and manifold settings of two texture descriptors show that combining different configurations always improves the accuracy compared to a predetermined configuration. On average, the CHMCS representation achieves 94.16% accuracy and outperforms deep learning networks and handcrafted color texture descriptors by over 5%, especially when the dataset is small. [ABSTRACT FROM AUTHOR]

The local binary model is a straightforward, dependable, and effective method for extracting relevant local information from images. However, because it only uses sign information in the local region, the local binary pattern (LBP) is ineffective at capturing discriminating characteristics. Furthermore, most LBP variants select a region with one specific center pixel to fill all neighborhoods. In this paper, a new variant of a LBP is proposed for texture classification, known as corner rhombus-shape LBP (CRSLBP). In the CRSLBP approach, we first use three methods to threshold the pixel's neighbors and center to obtain four center pixels by using sign and magnitude information with respect to a chosen region of an even block. This helps determine not just the relationship between neighbors and the pixel center but also between the center and the neighbor pixels of neighborhood center pixels. We evaluated the performance of our descriptors using four challenging texture databases: Outex (TC10,TC12), Brodatz, KTH-TIPSb2, and UMD. Various extensive experiments were performed that demonstrated the effectiveness and robustness of our descriptor in comparison with the available state of the art (SOTA). [ABSTRACT FROM AUTHOR]

Interest in automated data classification and identification systems has increased over the past years in conjunction with the high demand for artificial intelligence and security applications. In particular, recognizing human activities with accurate results have become a topic of high interest. Although the current tools have reached remarkable successes, it is still a challenging problem due to various uncontrolled environments and conditions. In this paper two statistical frameworks based on nonparametric hierarchical Bayesian models and Gamma distribution are proposed to solve some real-world applications. In particular, two nonparametric hierarchical Bayesian models based on Dirichlet process and Pitman-Yor process are developed. These models are then applied to address the problem of modelling grouped data where observations are organized into groups and these groups are statistically linked by sharing mixture components. The choice of the Gamma mixtures is motivated by its flexibility for modelling heavy-tailed distributions. In addition, deploying the Dirichlet process prior is justified by its advantage of automatically finding the right number of components and providing nice properties. Moreover, a learning step via variational Bayesian setting is presented in a flexible way. The priors over the parameters are selected appropriately and the posteriors are approximated effectively in a closed form. Experimental results based on a real-life applications that concerns texture classification and human actions recognition show the capabilities and effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]