Fuzzy Local Binary Pattern Research Articles

A Novel Statistical Feature Analysis-Based Global and Local Method for Face Recognition

Face recognition from an image/video has been a fast-growing area in research community, and a sizeable number of face recognition techniques based on texture analysis have been developed in the past few years. Further, these techniques work well on gray-scale and colored images, but very few techniques deal with binary and low-resolution images. As the binary image is becoming the preferred format for low face resolution analysis, there is a need for further studies to provide a complete solution for the image-based face recognition system with a higher accuracy rate. To overcome the limitation of the existing methods in extracting distinctive features in low-resolution images due to the contrast between the face and background, we propose a statistical feature analysis technique to fill the gaps. To achieve this, the proposed technique integrates the binary-level occurrence matrix (BLCM) and the fuzzy local binary pattern (FLBP) named FBLCM to extract global and local features of the face from binary and low-resolution images. The purpose of FBLCM is to distinctively improve performance of edge sharpness between black and white pixels in the binary image and to extract significant data relating to the features of the face pattern. Experimental results on Yale and FEI datasets validate the superiority of the proposed technique over the other top-performing feature analysis methods. The developed technique has achieved the accuracy of 94.54% when a random forest classifier is used, hence outperforming other techniques such as the gray-level co-occurrence matrix (GLCM), bag of word (BOW), and fuzzy local binary pattern (FLBP), respectively.

Improved background modeling of video sequences using spatio-temporal extension of fuzzy local binary pattern

Background subtraction is a method of motion detection in video sequences captured by static camera based on construction of a background model and its progressive comparison with each frame of the video. Sometimes the changes in the background objects are not permanent and appear at a rate faster than that of the background update, and this leads to emergence of dynamic textures in the background. As a result, high-quality background modeling plays a major role in motion detection performance, especially in videos with dynamic backgrounds and adverse environmental conditions such as noise. Although Local Binary Pattern (LBP) is a successful methodology for background subtraction, but it cannot properly extract textures from uniform areas of the foreground. In recent years, Fuzzy Local Binary Pattern (FLBP) has been developed to improve the performance of LBP operator in texture extraction from images with additive noise. The use of FLBP texture descriptor for background subtraction has led to improved robustness against noise and low sensitivity of the background model to slight changes in the texture gray scale values, and therefore better texture extraction from uniform areas, even in the presence of dynamic backgrounds and adverse environmental conditions. But despite this improvement, this operator is still sensitive to time-variations of pixels in dynamic backgrounds. To avoid this issue and incorporate correlation of pixel values over successive frames, this study proposes the use of spatio-temporal extension of FLBP with symmetry about central pixel in combination with the Local Color Histogram (LCH) in the Improved Hue Luminance and Saturation (IHLS) color space for describing the pixel color features. The results of tests conducted with standard databases show that for dynamic backgrounds, the use of proposed Spatio-Temporal Fuzzy Center Symmetric Local Binary Pattern (STFCS-LBP) operator with the spatio-temporal neighborhood texture patterns and the local color histogram yields better results than the existing methods.

Content-based Image Retrieval for Medicinal Plants Identification using Ensemble of SVM

In order to protect human health, it is very importance to know medicinal plant accurately. The plant identification is not easy and time-consuming task because of diversity. This paper proposes content-based image retrieval (CBIR) for medicinal plant identification using ensemble of Support Vector Machine (SVM). This research developed computerized geometric morphometric method to extract global shape descriptor. We investigated the combination of leaf texture and leaf geometry features based on Fuzzy Local Binary Pattern (FLBP) and global shape descriptor. Weighted Similarity-based Combination method was used to combine the features. We evaluated 660 medicinal leaves images of 66 species. The experiment results show that SVM ensemble of multiple features has good performance with the accuracy of classification is 0.93. CBIR system is developed for medicinal plant identification. The average precision of the CBIR system is 0.913. The proposed method is promising and potential to identify medicinal plant automatically.

Automated gait-based gender identification using fuzzy local binary patterns with tuned parameters

With the rapid advances in sensor networks and internet-of-things, several applications have evolved to ubiquitously monitor what is happening in smart environments. This paper designs a gait-based gender identification system based on an improved method of fuzzy local binary patterns for texture analysis. The moving person’s silhouette is extracted and a texture image is constructed to summarize structural and dynamical variations over one gait cycle. Then, histograms representing soft pattern variations are constructed and used as feature vectors. The proposed method employs grid search to find the optimal hyper-parameters for the feature extraction method. Finally, support vector machines are trained to predict the gender of the walking person. The performance is evaluated and compared with four other local binary pattern (LBP) related texture descriptors. The results show significant improvements in the performance under different walking conditions.

Intuitionistic fuzzy local binary pattern for features extraction

We propose a novel intuitionistic fuzzy feature extraction method to encode local texture. The proposed method extends the fuzzy local binary pattern approach by incorporating intuitionistic fuzzy set theory in the representation of local patterns of texture in images. Intuitionistic fuzzy local binary pattern also contribute to more than one bin in the distribution of the intuitionistic fuzzy local binary pattern values which can be used as a feature vector. The proposed intuitionistic fuzzy local binary pattern approach was experimentally evaluated for Lena image of the size 256 × 256. The results validate the effectiveness of the proposed intuitionistic fuzzy local binary pattern over the local binary pattern and fuzzy local binary pattern feature extraction methods.

Parallel Technique for Medicinal Plant Identification System using Fuzzy Local Binary Pattern

As biological image databases are growing rapidly, automated species identification based on digital data becomes of great interest for accelerating biodiversity assessment, research and monitoring. This research applied high performance computing (HPC) to a medicinal plant identification system. A parallel technique for medicinal plant image processing using Fuzzy Local Binary Pattern (FLBP) is proposed. The FLBP method extends the Local Binary Pattern (LBP) approach by employing fuzzy logic to represent texture images. The main goal of this research was to measure the efficiency of using the proposed parallel technique for medicinal plant image processing and evaluation in order to find out whether this approach is reasonable for handling large data sets. The parallel processing technique was designed in a message-sending model. 30 species of Indonesian medical plants were analyzed. Each species was represented by 48 leaf images. Performance evaluation was measured using the speed-up, efficiency, and isoefficiency of the parallel computing technique. Preliminary results show that HPC worked well in reducing the execution time of medical plant identification. In this work, parallel processing of training images was 7.64 times faster than with sequential processing, with efficiency values greater than 0.9. Parallel processing of testing images was 6.73 times faster than with sequential processing, with efficiency values over 0.9. The system was able to identify images with an accuracy of 68.89%.

CBIR of Brain MR Images Using Histogram of Fuzzy Oriented Gradients and Fuzzy Local Binary Patterns

Retrieval of similar images from large dataset of brain images across patients would help the experts in the decision diagnosis process of diseases. Generally used feature extraction methods are color, texture and shape. In medical images texture and shape features are most efficient. Histogram of Oriented Gradients (HOG) and Local Binary Pattern (LBP) are good descriptor for brain MR image retrieval. But there are many challenges facing in medical application. An empirical study of the impact of increasing bins number in the HOG descriptor concluded that larger the number is more accurate the descriptor is. In fact this is due to the reduction of orientations range that each bin covers. Despite the efficiency of augmenting the bins number, this technique has limited spatial support as the augmentation of the number of bins used leads to increase the histogram dimension. So here proposed a method called Histogram of Fuzzy Oriented Gradients (HFOG), in which a pixel can belong several bins with different degrees. The Local Binary Patterns feature extraction method is widely used for texture analysis; however, the original LBP is based on hard thresholding the neighborhood of each pixel. Therefore, texture representation with LBP is very sensitive to noise and cannot distinguish between a strong and a weak pattern. In this study, Fuzzy Local Binary Patterns was introduced to improve the original LBP.

CBIR of Brain MR Images Using Histogram of Fuzzy Oriented Gradients and Fuzzy Local Binary Patterns

CBIR of Brain MR Images Using Histogram of Fuzzy Oriented Gradients and Fuzzy Local Binary Patterns

Silhouette-Based Gender Recognition in Smart Environments Using Fuzzy Local Binary Patterns and Support Vector Machines

Abstract: Automatic recognition of human gender in smart environments is an interesting problem in biometric and demographic studies. In this paper, we describe a method for gender recognition at a distance based on visual texture analysis of gait energy images (GEIs). These images summarize the structural and dynamical variations of the subject's silhouette during one gait cycle. Texture analysis and feature extraction are based on histogram calculation of fuzzy local binary pattern (FLBP), which describes the relative intensities of each pixel with surrounding neighbors. Unlike the original LBP, each pixel can contribute, with different weights, to more than one bin in the histogram of occurring codes. The classification model uses support vector machines with linear kernel function. The performance of the proposed approach is intensively evaluated and compared with other texture on CASIA B multiview gait database. We also consider the variation of some conditions such as clothing and carried objects. Results show that the proposed approach is promising and outperforms other variants in representing texture for gait-based gender recognition.

Texture Feature Extraction Using Intuitionistic Fuzzy Local Binary Pattern

Abstract In this paper, intuitionistic fuzzy local binary for texture feature extraction (IFLBP) has been proposed to encode local texture from the input image. The proposed method extends the fuzzy local binary pattern approach by incorporating intuitionistic fuzzy sets in the representation of local patterns of texture in images. Intuitionistic fuzzy local binary pattern also contributes to more than one bin in the distribution of IFLBP values, which can further be used as a feature vector in the various fields of image processing. The performance of the proposed method has been demonstrated on various medical images and processing images of size 256×256. The obtained results validated the effectiveness and usefulness of our proposed method over the other reported methods, and new improvements are suggested.

Texture and fuzzy colour features to index Roman mosaic-images

We present an efficient approach to index and retrieve Roman mosaic-images by region. The image regions are defined based on the mosaic-image presentation. For more efficiency, the proposed approach uses an index combining texture and fuzzy colour features. These features are fuzzy image dominant colour, fuzzy colour histogram and local Robert gradient binary pattern operator LRGBP. Similar images are extracted using a fuzzy similarity and histogram intersection measures. Our proposed index is tested on database containing 800 images of historical Roman mosaics from the 1st to 4th century. Tests demonstrate that the fusion of the fuzzy colour and texture features increases the precision of the retrieval system. Evaluation tests and comparison with several other prevailing approaches prove a remarkable performance of our proposed approach.

Evaluation of noise robustness for local binary pattern descriptors in texture classification

Local binary pattern (LBP) operators have become commonly used texture descriptors in recent years. Several new LBP-based descriptors have been proposed, of which some aim at improving robustness to noise. To do this, the thresholding and encoding schemes used in the descriptors are modified. In this article, the robustness to noise for the eight following LBP-based descriptors are evaluated; improved LBP, median binary patterns (MBP), local ternary patterns (LTP), improved LTP (ILTP), local quinary patterns, robust LBP, and fuzzy LBP (FLBP). To put their performance into perspective they are compared to three well-known reference descriptors; the classic LBP, Gabor filter banks (GF), and standard descriptors derived from gray-level co-occurrence matrices. In addition, a roughly five times faster implementation of the FLBP descriptor is presented, and a new descriptor which we call shift LBP is introduced as an even faster approximation to the FLBP. The texture descriptors are compared and evaluated on six texture datasets; Brodatz, KTH-TIPS2b, Kylberg, Mondial Marmi, UIUC, and a Virus texture dataset. After optimizing all parameters for each dataset the descriptors are evaluated under increasing levels of additive Gaussian white noise. The discriminating power of the texture descriptors is assessed using tenfolded cross-validation of a nearest neighbor classifier. The results show that several of the descriptors perform well at low levels of noise while they all suffer, to different degrees, from higher levels of introduced noise. In our tests, ILTP and FLBP show an overall good performance on several datasets. The GF are often very noise robust compared to the LBP-family under moderate to high levels of noise but not necessarily the best descriptor under low levels of added noise. In our tests, MBP is neither a good texture descriptor nor stable to noise.

Fusion of fuzzy statistical distributions for classification of thyroid ultrasound patterns

This paper proposes a novel approach for thyroid ultrasound pattern representation. Considering that texture and echogenicity are correlated with thyroid malignancy, the proposed approach encodes these sonographic features via a noise-resistant representation. This representation is suitable for the discrimination of nodules of high malignancy risk from normal thyroid parenchyma. The material used in this study includes a total of 250 thyroid ultrasound patterns obtained from 75 patients in Greece. The patterns are represented by fused vectors of fuzzy features. Ultrasound texture is represented by fuzzy local binary patterns, whereas echogenicity is represented by fuzzy intensity histograms. The encoded thyroid ultrasound patterns are discriminated by support vector classifiers. The proposed approach was comprehensively evaluated using receiver operating characteristics (ROCs). The results show that the proposed fusion scheme outperforms previous thyroid ultrasound pattern representation methods proposed in the literature. The best classification accuracy was obtained with a polynomial kernel support vector machine, and reached 97.5% as estimated by the area under the ROC curve. The fusion of fuzzy local binary patterns and fuzzy grey-level histogram features is more effective than the state of the art approaches for the representation of thyroid ultrasound patterns and can be effectively utilized for the detection of nodules of high malignancy risk in the context of an intelligent medical system.