Principal curvatures based local binary pattern for rotation invariant texture classification

In this paper, an Extended Mapping Local Binary Pattern (EMLBP) method is proposed that is used for texture feature extraction. In this method, by extending nonuniform patterns a new mapping technique is suggested that extracts more discriminative features from textures. This new mapping is tested for some LBP operators such as CLBP, LBP, and LTP to improve the classification rate of them. The proposed approach is used for coding nonuniform patterns into more than one feature. The proposed method is rotation invariant and has all the positive points of previous approaches. By concatenating and joining two or more histograms significant improvement can be made for rotation invariant texture classification. The implementation of proposed mapping on Outex, UIUC and CUReT datasets shows that proposed method can improve the rate of classifications. Furthermore, the introduced mapping can increase the performance of any rotation invariant LBP, especially for large neighborhood. The most accurate result of the proposed technique has been obtained for CLBP. It is higher than that of some state-of-the-art LBP versions such as multiresolution CLBP and CLBC, DLBP, VZ_MR8, VZ_Joint, LTP, and LBPV.

Download Full-text

Comparative Analysis of Texture Classification Using Local Binary Pattern and Its Variants

International Journal of Information System Modeling and Design ◽

10.4018/ijismd.2017040103 ◽

2017 ◽

Vol 8 (2) ◽

pp. 45-56

Author(s):

Richa Sharma ◽

Madan Lal

Keyword(s):

Image Processing ◽

Comparative Analysis ◽

Digital Image Processing ◽

High Performance ◽

Local Binary Pattern ◽

Texture Classification ◽

Local Binary Patterns ◽

Classification Performance ◽

Powerful Method ◽

Rotation Invariant

Texture classification is an important issue in digital image processing and the Local Binary pattern (LBP) is a very powerful method used for analysing textures. LBP has gained significant popularity in texture analysis world. However, LBP method is very sensitive to noise and unable to capture the macrostructure information of the image. To address its limitation, some variants of LBP have been defined. In this chapter, the texture classification performance of LBP has been compared with the five latest high-performance LBP variants, like Centre symmetric Local Binary Pattern (CS-LBP), Orthogonal Combination of Local Binary Patterns (OC LBP), Rotation Invariant Local Binary Pattern (RLBP), Dominant Rotated Local Binary Pattern (DRLBP) and Median rotated extended local binary pattern (MRELBP). This was by using the standard images Outex_TC_0010 dataset. From the experimental results it is concluded that DRLBP and MRELBP are the best methods for texture classification.

Download Full-text

First- and Second-Order Sorted Local Binary Pattern Features for Grayscale-Inversion and Rotation Invariant Texture Classification

2020 25th International Conference on Pattern Recognition (ICPR) ◽

10.1109/icpr48806.2021.9412246 ◽

2021 ◽

Author(s):

Tiecheng Song ◽

Yuanjing Han ◽

Jie Feng ◽

Yuanlin Wang ◽

Chenqiang Gao

Keyword(s):

Local Binary Pattern ◽

Texture Classification ◽

Second Order ◽

Rotation Invariant

Download Full-text

Completed Grayscale-Inversion and Rotation Invariant Local Binary Pattern for Texture Classification

2018 24th International Conference on Pattern Recognition (ICPR) ◽

10.1109/icpr.2018.8545438 ◽

2018 ◽

Author(s):

Tiecheng Song ◽

Liangliang Xin ◽

Lin Luo ◽

Chenqiang Gao

Keyword(s):

Local Binary Pattern ◽

Texture Classification ◽

Rotation Invariant

Download Full-text

Completed Local Ternary Pattern for Rotation Invariant Texture Classification

The Scientific World JOURNAL ◽

10.1155/2014/373254 ◽

2014 ◽

Vol 2014 ◽

pp. 1-10 ◽

Cited By ~ 17

Author(s):

Taha H. Rassem ◽

Bee Ee Khoo

Keyword(s):

Classification Accuracy ◽

Local Binary Pattern ◽

Texture Classification ◽

Experimental Results ◽

Rotation Invariant ◽

Local Ternary Pattern ◽

Binary Count ◽

Robust To Noise

Despite the fact that the two texture descriptors, the completed modeling of Local Binary Pattern (CLBP) and the Completed Local Binary Count (CLBC), have achieved a remarkable accuracy for invariant rotation texture classification, they inherit some Local Binary Pattern (LBP) drawbacks. The LBP is sensitive to noise, and different patterns of LBP may be classified into the same class that reduces its discriminating property. Although, the Local Ternary Pattern (LTP) is proposed to be more robust to noise than LBP, however, the latter’s weakness may appear with the LTP as well as with LBP. In this paper, a novel completed modeling of the Local Ternary Pattern (LTP) operator is proposed to overcome both LBP drawbacks, and an associated completed Local Ternary Pattern (CLTP) scheme is developed for rotation invariant texture classification. The experimental results using four different texture databases show that the proposed CLTP achieved an impressive classification accuracy as compared to the CLBP and CLBC descriptors.

Download Full-text