Boosting Texture-Based Classification by Describing Statistical Information of Gray-Levels Differences
<p>Overview of Complete Local Oriented Statistical Information Booster (CLOSIB) method. Example about the calculation of CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics></math> on an image.</p> "> Figure 2
<p><math display="inline"><semantics> <msub> <mo>Δ</mo> <mi>p</mi> </msub> </semantics></math> images showing the absolute differences of the gray values for <math display="inline"><semantics> <mrow> <mi>P</mi> <mo>=</mo> <mn>8</mn> </mrow> </semantics></math> orientations in a neighborhood of radii <math display="inline"><semantics> <mrow> <mi>R</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>R</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math>. The original image <span class="html-italic">I</span> is shown in the center. The main change in the intensity of the original image occurs in the horizontal direction <math display="inline"><semantics> <mrow> <mi>p</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>p</mi> <mo>=</mo> <mn>5</mn> </mrow> </semantics></math>.</p> "> Figure 3
<p>(Better viewed in color) Neighborhood around a center pixel, (28), considered for the computation of M-CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mi>θ</mi> <mo>∥</mo> <mn>8</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mi>θ</mi> <mo>∥</mo> <mn>8</mn> <mo>,</mo> <mn>3</mn> <mo>,</mo> <mi>θ</mi> </mrow> </msub> </semantics></math> and HM-CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mi>θ</mi> <mo>∥</mo> <mn>8</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mi>θ</mi> <mo>∥</mo> <mn>8</mn> <mo>,</mo> <mn>3</mn> <mo>,</mo> <mi>θ</mi> </mrow> </msub> </semantics></math>. CLOSIB considers <span class="html-italic">P</span> = 8 orientations, while H-CLOSIB only <span class="html-italic">P</span> = 4 orientations. In the figure, neighbour pixels considered for H-CLOSIB are shown in bold.</p> "> Figure 4
<p>(<b>a</b>) Circumference that represents the neighborhood considered for the computation of CLOSIB with <math display="inline"><semantics> <mrow> <mi>P</mi> <mo>=</mo> <mn>8</mn> </mrow> </semantics></math>. Four pairs of neighbors differ in <math display="inline"><semantics> <mi>π</mi> </semantics></math> radians, such as the neighbors for values <math display="inline"><semantics> <mrow> <mi>p</mi> <mo>=</mo> <mn>2</mn> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>p</mi> <mo>=</mo> <mn>6</mn> </mrow> </semantics></math>. (<b>b</b>,<b>c</b>) Schemas that represent the computation of CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mi>θ</mi> </mrow> </msub> </semantics></math> for two different images. We show the original image in the centre and the eight images of the absolute differences of the gray values <math display="inline"><semantics> <mrow> <msub> <mi>d</mi> <mi>p</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>c</mi> </msub> <mo>,</mo> <msub> <mi>y</mi> <mi>c</mi> </msub> <mo>)</mo> </mrow> </mrow> </semantics></math> in the outer layer. The red and green numbers indicate the values of each element of CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics></math> and CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> </mrow> </msub> </semantics></math> feature set, respectively, obtained for the corresponding <span class="html-italic">p</span> values of <math display="inline"><semantics> <mrow> <msub> <mi>d</mi> <mi>p</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>c</mi> </msub> <mo>,</mo> <msub> <mi>y</mi> <mi>c</mi> </msub> <mo>)</mo> </mrow> </mrow> </semantics></math>. Note that the values of the elements of CLOSIB computed for neighbors that differ in <math display="inline"><semantics> <mi>π</mi> </semantics></math> radians diverge in only a maximum of 0.0002 units whereas the ones that differ in a different angle diverge in at least 0.0006 units.</p> "> Figure 5
<p>Schemas of the computation of CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mi>θ</mi> </mrow> </msub> </semantics></math> (<b>left</b>) and H-CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mi>θ</mi> </mrow> </msub> </semantics></math> (<b>right</b>) using the example of <a href="#sensors-19-01048-f004" class="html-fig">Figure 4</a>c.</p> "> Figure 6
<p>(<b>a</b>) Example of gray values of an Original Image. (<b>b</b>) Matrix obtained from the first difference of the gray values at 0 degrees. It corresponds with the first element of CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics></math>. (<b>d</b>) Matrix obtained from the first difference of the gray values at 180 degrees. It corresponds with the fifth element of CLOSIB<math display="inline"><semantics> <msub> <mrow/> <mrow> <mn>8</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics></math>. (<b>c</b>) Differences, with sign, at 0 degrees. CLOSIB uses the absolute value of the differences, therefore these values with sign are never computed.</p> "> Figure 7
<p>Hit rates when we describe KTH Tips2-a images with different CLOSIBs and LBP-based descriptors. For each CLOSIB variant –CLOSIB (standard), M-CLOSIB, H-CLOSIB and HM-CLOSIB–, we only represent the best result obtained among the results with different combinations of parameters.</p> "> Figure 8
<p>Hit rates obtained with a given LBP-based descriptor (LBP, ALBP, LBPV and CLBP) and the concatenations of the descriptor with CLOSIB variants.</p> "> Figure 9
<p>Hit rates for LBP-based descriptors LBP, ALBP, CLBP and LBPV and their multi-scale versions.</p> "> Figure 10
<p>Hit rates obtained with the concatenation of multi-scale LBP-based descriptors and HM-CLOSIB. The horizontal line represents the hit rate of HM-CLOSIB descriptor.</p> "> Figure 11
<p>Results using the concatenation of LBP-based descriptors with CLOSIB variants (CLOSIB, H-CLOSIB, M-CLOSIB and HM-CLOSIB) on UIUC (<b>left</b>) and USPTex (<b>right</b>) dataset.</p> "> Figure 12
<p>Results using the concatenation of LBP-based descriptors with CLOSIB variants (CLOSIB, H-CLOSIB, M-CLOSIB and HM-CLOSIB) on the original images (<b>left</b>) and the cropped ones (<b>right</b>) of JAFFE dataset.</p> ">
Abstract
:1. Introduction
2. Related Works
2.1. Descriptors Based on LBP
2.1.1. Local Binary Patterns (LBP)
2.1.2. Adaptive Local Binary Patterns (ALBP)
2.1.3. Local Binary Patterns Variance (LBPV)
2.1.4. Completed Local Binary Patterns (CLBP)
3. Method
3.1. Overview
3.2. Complete Local Oriented Statistical Information Booster (CLOSIB)
3.3. CLOSIB Variants
3.3.1. Multi-Scale CLOSIB (M-CLOSIB)
3.3.2. Half CLOSIB (H-CLOSIB)
3.3.3. Half Multi-Scale CLOSIB (HM-CLOSIB)
4. Experiments and Results
4.1. Datasets
4.1.1. KTH TIPS2-a
4.1.2. UIUC
4.1.3. USPTex
4.1.4. JAFFE
4.2. Experimental Setup
4.3. Results for KTH Tips2-a
4.3.1. CLOSIB versus LBP-Based Descriptors
4.3.2. CLOSIB and LBP-Based Descriptors
4.3.3. No Multi-Scale versus Multi-Scale LBP-Based Descriptors
4.3.4. HM-CLOSIB + Multi-Scale LBP-Based Descriptors
4.3.5. Comparative with the State-of-the-Art
4.4. Results for UIUC and USPTex
4.5. Results for JAFFE
4.6. Computational Cost of CLOSIB and LBP Variants
5. Conclusions
Author Contributions
Funding
Acknowledgments
Conflicts of Interest
Abbreviations
LBP | Local Binary Pattern |
ALBP | Adaptive Local Binary Pattern |
ALBPV | Adaptive Local Binary Pattern Variance |
CLOSIB | Complete Local Oriented Statistical Information Booster |
H-CLOSIB | Half Complete Local Oriented Statistical Information Booster |
M-CLOSIB | Multi-scale Complete Local Oriented Statistical Information Booster |
ASASEC | Advisory System Against Sexual Exploitation of Children |
CDC | Compact Digital Cameras |
CNN | Convolutional Neural Networks |
CSA | Child Sexual Abuse |
References
- Liu, J.; He, J.; Zhang, W.; Xu, P.; Tang, Z. TCvBsISM: Texture Classification via B-Splines-Based Image Statistical Modeling. IEEE Access 2018, 6, 44876–44893. [Google Scholar] [CrossRef]
- Hossain, S.; Serikawa, S. Texture databases—A comprehensive survey. Pattern Recognit. Lett. 2013, 34, 2007–2022. [Google Scholar] [CrossRef]
- Bianconi, F.; Fernández, A. An appendix to “Texture databases—A comprehensive survey”. Pattern Recognit. Lett. 2014, 45, 33–38. [Google Scholar] [CrossRef]
- Backes, A.R.; de Mesquita Sá Junior, J.J. LBP maps for improving fractal based texture classification. Neurocomputing 2017, 266, 1–7. [Google Scholar] [CrossRef]
- Florindo, J.B.; Landini, G.; Bruno, O.M. Three-dimensional connectivity index for texture recognition. Pattern Recognit. Lett. 2016, 84, 239–244. [Google Scholar] [CrossRef] [Green Version]
- Cernadas, E.; Fernández-Delgado, M.; González-Rufino, E.; Carrión, P. Influence of normalization and color space to color texture classification. Pattern Recognit. 2017, 61, 120–138. [Google Scholar] [CrossRef]
- Casanova, D.; Florindo, J.; Falvo, M.; Bruno, O. Texture analysis using fractal descriptors estimated by the mutual interference of color channels. Inf. Sci. 2016, 346, 58–72. [Google Scholar] [CrossRef]
- Dutta, S.; Pal, S.K.; Sen, R. On-machine tool prediction of flank wear from machined surface images using texture analyses and support vector regression. Precis. Eng. 2016, 43, 34–42. [Google Scholar] [CrossRef]
- Castrillón-Santana, M.; Lorenzo-Navarro, J.; Ramón-Balmaseda, E. Fusion of Holistic and Part Based Features for Gender Classification in the Wild. In New Trends in Image Analysis and Processing–ICIAP 2015 Workshops; Murino, V., Puppo, E., Sona, D., Cristani, M., Sansone, C., Eds.; Springer International Publishing: New York, NY, USA, 2015; Volume 9281, pp. 43–50. [Google Scholar]
- Kavitha, M.S.; An, S.Y.; An, C.H.; Huh, K.H.; Yi, W.J.; Heo, M.S.; Lee, S.S.; Choi, S.C. Texture analysis of mandibular cortical bone on digital dental panoramic radiographs for the diagnosis of osteoporosis in Korean women. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. 2015, 119, 346–356. [Google Scholar] [CrossRef] [PubMed]
- Lan, Z.; Liu, Y. Study on Multi-Scale Window Determination for GLCM Texture Description in High-Resolution Remote Sensing Image Geo-Analysis Supported by GIS and Domain Knowledge. Int. J. Geo-Inf. 2018, 7, 175. [Google Scholar] [CrossRef]
- González-Castro, V.; Alegre, E.; García-Olalla, O.; Fernández-Robles, L.; García-Ordás, M.T. Adaptive pattern spectrum image description using Euclidean and Geodesic distance without training for texture classification. IET Comput. Vis. 2012, 6, 581–589. [Google Scholar] [CrossRef]
- Alegre, E.; Barreiro, J.; Suárez-Castrillón, S. A new improved Laws-based descriptor for surface roughness evaluation. Int. J. Adv. Manuf. Technol. 2012, 59, 605–615. [Google Scholar] [CrossRef]
- Mallikarjuna, P.; Targhi, A.T.; Fritz, M.; Hayman, E.; Caputo, B.; Eklundh, J.O. THE KTH-TIPS2 Database. 2006. Available online: http://www.nada.kth.se/cvap/databases/kth-tips/kth-tips2.pdf (accessed on 28 February 2019).
- Caputo, B.; Hayman, E.; Mallikarjuna, P. Class-specific material categorisation. In Proceedings of the 10th IEEE International Conference on Computer Vision (ICCV’05), Beijing, China, 17–20 October 2005; pp. 1597–1604. [Google Scholar]
- Chen, J.; Shan, S.; He, C.; Zhao, G.; Pietikainen, M.; Member, S.; Chen, X.; Member, S.; Gao, W. WLD: A robust local image descriptor. IEEE Trans. Pattern Anal. Mach. Intell. 2010, 32, 1705–1720. [Google Scholar] [CrossRef] [PubMed]
- Hussain, S.; Triggs, B. Visual Recognition Using Local Quantized Patterns. In Computer Vision ECCV 2012; Lecture Notes in Computer Science; Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C., Eds.; Springer: Berlin/Heidelberg, Germany, 2012; pp. 716–729. [Google Scholar]
- Hafiane, A.; Palaniappan, K.; Seetharaman, G. Joint Adaptive Median Binary Patterns for texture classification. Pattern Recognit. 2015, 48, 2609–2620. [Google Scholar] [CrossRef]
- Xiang, Z.; Tan, H.; Ye, W. The Excellent Properties of a Dense Grid-Based HOG Feature on Face Recognition Compared to Gabor and LBP. IEEE Access 2018, 6, 29306–29318. [Google Scholar] [CrossRef]
- Elaiwat, S.; Bennamoun, M.; Boussaid, F.; El-Sallam, A. A Curvelet-based approach for textured 3D face recognition. Pattern Recognit. 2015, 48, 1235–1246. [Google Scholar] [CrossRef]
- De Marsico, M.; Nappi, M.; Riccio, D.; Wechsler, H. Robust face recognition after plastic surgery using region-based approaches. Pattern Recognit. 2015, 48, 1261–1276. [Google Scholar] [CrossRef]
- Lyons, M.; Akamatsu, S.; Kamachi, M.; Gyoba, J. Coding facial expressions with Gabor wavelets. In Proceedings of the Third IEEE International Conference on Automatic Face and Gesture Recognition, Nara, Japan, 14–16 April 1998; pp. 200–205. [Google Scholar]
- Yang, B.; Cao, J.; Ni, R.; Zhang, Y. Facial Expression Recognition using Weighted Mixture Deep Neural Network Based on Double-channel Facial Images. IEEE Access 2017. [Google Scholar] [CrossRef]
- Shenoy, P.D.; Iyengar, S.S.; Raja, K.B.; Buyya, R.; Patnaik, L.M.; Rangaswamy, Y.; Raja, K.; Venugopal, K. FRDF: Face Recognition Using Fusion of DTCWT and FFT Features. Procedia Comput. Sci. 2015, 54, 809–817. [Google Scholar] [Green Version]
- Wan, W.; Zhou, Z.; Zhao, J.; Cao, F. A novel face recognition method: Using random weight networks and quasi-singular value decomposition. Neurocomputing 2015, 151 Pt 3, 1180–1186. [Google Scholar] [CrossRef]
- Zang, F.; Zhang, J.; Pan, J. Face recognition using Elasticfaces. Pattern Recognit. 2012, 45, 3866–3876. [Google Scholar] [CrossRef]
- Ojala, T.; Pietikainen, M.; Harwood, D. Performance evaluation of texture measures with classification based on Kullback discrimination of distributions. In Proceedings of the 12th International Conference on Pattern Recognition, Jerusalem, Israel, 9–13 October 1994; Volume 1, pp. 582–585. [Google Scholar]
- Ahonen, T.; Matas, J.; He, C.; Pietikainen, M. Rotation Invariant Image Description with Local Binary Pattern Histogram Fourier Features. In Image Analysis; Salberg, A.B., Hardeberg, J.Y., Jenssen, R., Eds.; Springer: Berlin/Heidelberg, Germany, 2009; pp. 61–70. [Google Scholar]
- Zhao, G.; Pietikainen, M. Dynamic texture recognition using local binary patterns with an application to facial expressions. IEEE Trans. Pattern Anal. Mach. Intell. 2007, 29, 915–928. [Google Scholar] [CrossRef] [PubMed]
- Guo, Y.; Zhao, G.; Pietikainen, M. Texture Classification using a Linear Configuration Model based Descriptor. In Proceedings of the British Machine Vision Conference, Dundee, UK, 29 August–2 September 2011. [Google Scholar]
- Guo, Z.; Zhang, L.; Zhang, D.; Zhang, S. Rotation invariant texture classification using adaptive LBP with directional statistical features. In Proceedings of the 2010 17th IEEE International Conference on Image Processing (ICIP), Hong Kong, China, 26–29 September 2010; pp. 285–288. [Google Scholar]
- Guo, Z.; Zhang, L.; Zhang, D. Rotation invariant texture classification using LBP variance (LBPV) with global matching. Pattern Recognit. 2010, 43, 706–719. [Google Scholar] [CrossRef]
- Guo, Z.; Zhang, L.; Zang, D. A Completed Modeling of Local Binary Pattern Operator for Texture Classification. IEEE Trans. Image Process. 2010, 19, 1657–1663. [Google Scholar] [PubMed] [Green Version]
- Fernández, A.; Álvarez, M.X.; Bianconi, F. Texture Description Through Histograms of Equivalent Patterns. J. Math. Imaging Vis. 2013, 45, 76–102. [Google Scholar] [CrossRef]
- Liu, L.; Fieguth, P.; Guo, Y.; Wang, X.; Pietikäinen, M. Local binary features for texture classification: Taxonomy and experimental study. Pattern Recognit. 2017, 62, 135–160. [Google Scholar] [CrossRef]
- García-Olalla, O.; Alegre, E.; Fernández-Robles, L.; García-Ordás, M.T. Vitality assessment of boar sperm using an adaptive LBP based on oriented deviation. In Computer Vision—ACCV 2012 Workshops; Park, J.I., Kim, J., Eds.; Springer: Berlin/Heidelberg, Germany, 2013; pp. 61–72. [Google Scholar]
- García-Olalla, O.; Alegre, E.; Fernández-Robles, L.; García-Ordás, M.T.; García-Ordás, D. Adaptive Local Binary Pattern with oriented standard deviation (ALBPS) for texture classification. EURASIP J. Image Video Process. 2013, 2013, 31. [Google Scholar] [CrossRef]
- García-Olalla, O.; Alegre, E.; García-Ordás, M.T.; Fernández-Robles, L. Evaluation of LBP Variants using several Metrics and kNN Classifiers. In Similarity Search and Applications; Brisaboa, N., Pedreira, O., Zezula, P., Eds.; Springer: Berlin/Heidelberg, Germany, 2013; pp. 151–162. [Google Scholar]
- García-Olalla, O.; Alegre, E.; Fernández-Robles, L.; González-Castro, V. Local Oriented Statistics Information Booster (LOSIB) for Texture Classification. In Proceedings of the 2014 22nd International Conference on Pattern Recognition (ICPR), Stockholm, Sweden, 24–28 August 2014; pp. 1114–1119. [Google Scholar]
- García-Olalla Olivera, O. Methods for Improving Texture Description by Using Statistical Information Extracted from the Image Gradient. Ph.D. Thesis, Universidad de León, León, Spain, 2017. [Google Scholar]
- Lazebnik, S.; Schmid, C.; Ponce, J. A sparse texture representation using local affine regions. IEEE Trans. Pattern Anal. Mach. Intell. 2005, 27, 1265–1278. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Backes, A.R.; Casanova, D.; Bruno, O.M. Color texture analysis based on fractal descriptors. Pattern Recognit. 2012, 45, 1984–1992. [Google Scholar] [CrossRef] [Green Version]
- García-Olalla, O.; Alegre, E.; Fernández-Robles, L.; Fidalgo, E.; Saikia, S. Textile retrieval based on image content from CDC and webcam cameras in indoor environments. Sensors 2018, 18, 1329. [Google Scholar] [CrossRef] [PubMed]
- Ojala, T.; Pietikainen, M.; Maenpaa, T. Multiresolution gray-scale and rotation invariant texture classification with Local Binary Patterns. IEEE Trans. Pattern Anal. Mach. Intell. 2002, 24, 971–987. [Google Scholar] [CrossRef]
- Chan, C.H.; Kittler, J.; Messer, K. Multi-scale Local Binary Pattern Histograms for Face Recognition. In Advances in Biometrics; Lee, S.W., Li, S.Z., Eds.; Springer: Berlin/Heidelberg, Germany, 2007; pp. 809–818. [Google Scholar]
- Jia, X.; Yang, X.; Cao, K.; Zang, Y.; Zhang, N.; Dai, R.; Zhu, X.; Tian, J. Multi-scale local binary pattern with filters for spoof fingerprint detection. Inf. Sci. 2014, 268, 91–102. [Google Scholar] [CrossRef]
- Wen, Z.; Li, Z.; Peng, Y.; Ying, S. Virus image classification using multi-scale completed local binary pattern features extracted from filtered images by multi-scale principal component analysis. Pattern Recognit. Lett. 2016, 79, 25–30. [Google Scholar] [CrossRef]
- Hsu, C.W.; Lin, C.J. A Comparison of Methods for Multiclass Support Vector Machines. IEEE Trans. Neural Netw. 2002, 13, 415–425. [Google Scholar] [PubMed]
- Sharma, G.; ul Hussain, S.; Jurie, F. Local Higher-Order Statistics (LHS) for Texture Categorization and Facial Analysis. In Computer Vision ECCV 2012; Lecture Notes in Computer Science; Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C., Eds.; Springer: Berlin/Heidelberg, Germany, 2012; Volume 7578, pp. 1–12. [Google Scholar]
- Zhang, L.; Chu, R.; Xiang, S.; Liao, S.; Li, S.Z. Face Detection Based on Multi-Block LBP Representation. In Advances in Biometrics; Lee, S.W., Li, S.Z., Eds.; Springer: Berlin/Heidelberg, Germany, 2007; pp. 11–18. [Google Scholar]
- Viola, P.; Jones, M. Rapid object detection using a boosted cascade of simple features. In Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Kauai, HI, USA, 8–14 December 2001; Volume 1. [Google Scholar]
- Song, Y.; Zhang, F.; Li, Q.; Huang, H.; Odonnell, L.J.; Cai, W. Locally-Transferred Fisher Vectors for Texture Classification. In Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy, 22–29 October 2017; pp. 4922–4930. [Google Scholar]
- Vu, N.S.; Nguyen, V.L.; Gosselin, P.H. A Handcrafted Normalized-Convolution Network for Texture Classification. In Proceedings of the 2017 IEEE International Conference on Computer Vision Workshops, ICCVW 2017, Venice, Italy, 22–29 October 2017; pp. 1238–1245. [Google Scholar]
- Cimpoi, M.; Maji, S.; Kokkinos, I.; Mohamed, S.; Vedaldi, A. Describing Textures in the Wild. In Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 23–28 June 2014; pp. 3606–3613. [Google Scholar]
- Liu, F.; Tang, Z.; Tang, J. WLBP: Weber local binary pattern for local image description. Neurocomputing 2013, 120, 325–335. [Google Scholar] [CrossRef]
- Li, W.; Fritz, M. Recognizing Materials from Virtual Examples. In Proceedings of the 12th European Conference on Computer Vision—Volume Part IV; Springer: Berlin/Heidelberg, Germany, 2012; pp. 345–358. [Google Scholar]
- Zhang, J.; Zhao, H.; Liang, J. Continuous rotation invariant local descriptors for texton dictionary-based texture classification. Comput. Vis. Image Underst. 2013, 117, 56–75. [Google Scholar] [CrossRef]
- Satpathy, A.; Jiang, X.; Eng, H.L. LBP-Based Edge-Texture Features for Object Recognition. IEEE Trans. Image Process. 2014, 23, 1953–1964. [Google Scholar] [CrossRef] [PubMed]
- Voravuthikunchai, W.; Crémilleux, B.; Jurie, F. Histograms of pattern sets for image classification and object recognition. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 23–28 June 2014; pp. 224–231. [Google Scholar]
- Li, W. Learning Multi-scale Representations for Material Classification. In Pattern Recognition; Jiang, X., Hornegger, J., Koch, R., Eds.; Springer International Publishing: Cham, Switzerland, 2014; pp. 757–764. [Google Scholar]
- Wang, K.; Bichot, C.E.; Li, Y.; Li, B. Local Binary Circumferential and radial derivative pattern for texture classification. Pattern Recognit. 2017, 67, 213–229. [Google Scholar] [CrossRef]
- Gangwar, A.; Fidalgo, E.; Alegre, E.; González-Castro, V. Pornography and child sexual abuse detection in image and video: A comparative evaluation. In Proceedings of the 8th International Conference on Imaging for Crime Detection and Prevention (ICDP 2017), Madrid, Spain, 13–15 December 2017; pp. 37–42. [Google Scholar]
Parameter | Meaning |
---|---|
Gray value of the central pixel | |
Gray value of neighbor p | |
P | Number of neighbors |
R | Radius of the neighborhood |
Weight element used to minimize the directional difference | |
w | Weight, it is a constant between 0 to the maximum gray level value difference |
N | Number of rows in the image |
M | Number of columns in the image |
k | A bin of a histogram |
K | Maximum value of LBP |
u | Mean over the neighbors |
c | Threshold, mean value of the differences between the central pixel and neighbors |
Parameter | Meaning |
---|---|
I | Image |
c | Central pixel |
p | Neighbor pixel |
Gray value of pixel c | |
Gray value of neighbor pixel p | |
R | Radius of the neighborhood |
Absolute difference image at bearing p | |
moment of image | |
‖ | Concatenation function |
Order of the statistical moment considered | |
Variable that allows choosing between CLOSIB and H-CLOSIB |
Radius (R) | Neighbors (Orientations) (P) | Order () |
---|---|---|
1 | 8 | 1 |
1 | 8 | 2 |
2 | 16 | 1 |
2 | 16 | 2 |
1 | 8 | 1,2 |
2 | 16 | 1,2 |
Radius (R) | Neighbors (Orientations) (P) | Order () |
---|---|---|
1,2,3 | 8 | 1 |
1,2,3,4,5 | 8 | 1 |
1,2,3 | 8 | 2 |
1,2,3,4,5 | 8 | 2 |
2,3,4 | 16 | 1 |
2,3,4,5,6 | 16 | 1 |
2,3,4 | 16 | 2 |
2,3,4,5,6 | 16 | 2 |
1,2,3 | 8 | 1,2 |
1,2,3,4,5 | 8 | 1,2 |
2,3,4 | 16 | 1,2 |
2,3,4,5,6 | 16 | 1,2 |
Descriptor (D) | D | DC | DH-C | DM-C | DHM-C |
---|---|---|---|---|---|
LBP | 60.71 | 68.20 | 67.80 | 71.95 | 71.86 |
LBP | 65.53 | 70.52 | 70.33 | 71.78 | 72.50 |
ALBP | 56.46 | 64.86 | 64.84 | 68.97 | 69.15 |
ALBP | 65.97 | 68.96 | 68.88 | 69.84 | 70.16 |
LBPV | 59.30 | 67.05 | 67.51 | 69.89 | 71.15 |
LBPV | 62.27 | 69.00 | 69.24 | 70.14 | 71.17 |
CLBP | 64.37 | 69.63 | 69.95 | 71.97 | 71.76 |
CLBP | 67.53 | 71.95 | 72.54 | 72.01 | 72.54 |
Descriptor—Handcrafted | Hit Rate (%) | Reference |
---|---|---|
WLD | 56.4 | [16] |
MWLD | 64.7 | [16] |
SIFT | 52.7 | [16] |
LTP | 60.7 | [17] |
LQP | 64.2 | [17] |
WLBP | 64.4 | [55] |
LHS | 73.0 | [49] |
CMLBP | 73.1 | [56] |
CMR | 69.4 | [57] |
PC | 71.5 | [57] |
DRLTP | 62.6 | [58] |
DRLBP | 59.0 | [58] |
HoPS | 75.0 | [59] |
IFV | 82.2 | [54] |
AMBP | 70.3 | [18] |
MS4C | 70.5 | [60] |
CRDP(NNC) | 73.8 | [61] |
CRDP(SVM) | 78.0 | [61] |
HM-CLOSIB | 67.9 | Ours |
CLBPHM-CLOSIB | 74.8 | Ours |
Descriptor—Deep Features | Hit Rate (%) | Reference |
DeCAF | 78.4 | [54] |
LFV + FC-CNN | 82.6 | [52] |
NmzNet | 82.4 | [53] |
Dataset | LBP | LBP | ALBP | ALBP | LBPV | LBPV | CLBP | CLBP |
---|---|---|---|---|---|---|---|---|
UIUC | 0.09183 | 0.17636 | 0.20309 | 0,44195 | 0.15119 | 0.34011 | 0.1157 | 0.23574 |
USPTex | 0.00351 | 0.00488 | 0.00605 | 0.01051 | 0.00428 | 0.00918 | 0.00361 | 0.00594 |
KTH-TIPS2-a | 0.00914 | 0.01084 | 0.01297 | 0.02555 | 0.0114 | 0.02283 | 0.00754 | 0.01326 |
JAFFE | 0.01116 | 0.01695 | 0.02179 | 0.04263 | 0.01628 | 0.03312 | 0.01013 | 0.0183 |
Dataset | C | C | H-C | H-C | M-C | M-C | HM-C | HM-C |
---|---|---|---|---|---|---|---|---|
UIUC | 0.08655 | 0.18975 | 0.08630 | 0.18975 | 0.25392 | 0.59675 | 0.25142 | 0.56396 |
USPTex | 0.00393 | 0.00594 | 0.00377 | 0.00564 | 0.00938 | 0.01584 | 0.00912 | 0.01471 |
KTH TIPS2-a | 0.00921 | 0.01782 | 0.00838 | 0.01682 | 0.02284 | 0.05158 | 0.02266 | 0.04846 |
JAFFE | 0.02640 | 0.02790 | 0.01442 | 0.02594 | 0.03833 | 0.07177 | 0.03582 | 0.06767 |
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
García-Olalla, Ó.; Fernández-Robles, L.; Alegre, E.; Castejón-Limas, M.; Fidalgo, E. Boosting Texture-Based Classification by Describing Statistical Information of Gray-Levels Differences. Sensors 2019, 19, 1048. https://doi.org/10.3390/s19051048
García-Olalla Ó, Fernández-Robles L, Alegre E, Castejón-Limas M, Fidalgo E. Boosting Texture-Based Classification by Describing Statistical Information of Gray-Levels Differences. Sensors. 2019; 19(5):1048. https://doi.org/10.3390/s19051048
Chicago/Turabian StyleGarcía-Olalla, Óscar, Laura Fernández-Robles, Enrique Alegre, Manuel Castejón-Limas, and Eduardo Fidalgo. 2019. "Boosting Texture-Based Classification by Describing Statistical Information of Gray-Levels Differences" Sensors 19, no. 5: 1048. https://doi.org/10.3390/s19051048
APA StyleGarcía-Olalla, Ó., Fernández-Robles, L., Alegre, E., Castejón-Limas, M., & Fidalgo, E. (2019). Boosting Texture-Based Classification by Describing Statistical Information of Gray-Levels Differences. Sensors, 19(5), 1048. https://doi.org/10.3390/s19051048