A Hybrid Method of Textile Defect Detection using GLCM, LBP, SVD and Wavelet Transform
Chetan Chaudhari1, Ravindra Kumar Gupta2, Sapana Fegade3

1Mr. Chetan Chaudhari, Phd, Computer Science Engineering, RKDF Institute of Science & Technology, SRK university, Bhopal, MP.
2Dr. Ravindra Kumar Gupta, Associate Professor,Computer Science and Engineering Department, RKDFIST, BHOPAL, and M.P. India.
3Miss. Sapana Fegade, Assistant Professor, Computer Science Engineering , JTMCOE.
Manuscript received on February 28, 2020. | Revised Manuscript received on March 22, 2020. | Manuscript published on March 30, 2020. | PP: 5356-5360 | Volume-8 Issue-6, March 2020. | Retrieval Number: F9569038620/2020©BEIESP | DOI: 10.35940/ijrte.F9569.038620

Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: A roll of fabric with defects can have a depreciation of 45 to 65% with respect to the original price. While some commercial solutions exist, automatic fabric defect detection remains an active field of development and research. The goal is to extract the characteristics of the texture of the fabric to detect defects contained using image processing techniques. To date, there is no standard method which ensures the detection of texture defects in fabrics with high precision. In the following work, the use of Singular Value Decomposition (SVD), Local Binary Pattern (LBP) and Gray-Level Co-Occurrence Matrix (GLCM) features of images for the identification of defects in textiles is presented, where the application of techniques for pre-processing is presented, and for the analysis of texture LBP and the GLCM in order to extract features and segmentation is done using SVD approach. This model makes it possible to obtain compact and precise detection of the faulty texture structures. Our method is capable of achieving very precise detection and localization of texture defects in the images of the Fabric-Defect-Inspection-GLSR database, while ensuring a reasonable processing time.
Keywords: GLCM, LBP, SVD, and Wavelet transform.
Scope of the Article: Probabilistic Models and Methods.