Fabrics Fault Detection Techniques: A Survey

Abstract

Defect detection is an essential factor in the quality control process in textile manufacturing. Investments in the automated detection of texture defects make the cost of labour more economical. The fabric price is often influenced by defect-free fabrics, which draw a lot of attention in the textile industry. There are issues with manual inspections to ensure quality control, since they lack precision and consume many man hours. Therefore, these drawbacks need to be done away with by some automatic fabric inspection i.e. machine-vision-based inspection. To build an automatic inspection system, robust and powerful Fabric Defect Detection (FDD) techniques are required. Over the last few years, several FDD techniques using data mining with image processing algorithms have been proposed by many researchers. In this article, a detailed survey is presented on fabrics defect detection techniques using different data mining algorithms including image pre-processing, segmentation and feature extraction phases. It also discusses the benefits and disadvantages of the surveyed techniques used for detecting fabric defects. Eventually, a full review is completed and a few techniques are suggested to enhance fabric quality in potential directions.