Fabric defect classification using transfer learning and deep learning

Abstract

The internal inspection of fabrics is one of the most important phases of production in order to achieve high quality standard in the textile industry. Therefore, developing efficient automatic internal control mechanism has been an extremely major area of research. In this paper, the famous architecture Googlenet was fine-tuned into two configurations for texture defect classification that was trained on a textile texture database (TILDA). The experimental result, for both configurations, achieved a significant overall accuracy score of 97% for motif and a non-motif-based images and 89% for mixed images. In the results obtained, it was observed that the second model, which updates the last six layers, was more successful than the first one; which updates the last two layers.

Image classification-based transfer learning framework for image detection of IoT devices

“…The studies in [19]- [22] also apply CNN-based TL techniques to identify leukocytes as well as red blood cells for blood-related diseases and classify fundus for general retinal diseases diagnosis, respectively. The TL with pre-trained CNN models are also used for non-medical applications including Thai culture and Pitha traditional food images classification [23], [24], and other applications including land cover, fabric defect, birds' species, distracted driver classifications as well as age-invariant face recognition [25]- [29]. All the above literature studies show that the TL approach can achieve better accuracy performance than those of non-TL methods for various considered datasets and scenarios.…”