FABRIC CLASSIFICATION USING OPENCV2

This project presents a comprehensive and automated fabric classification system developed using OpenCV2 and machine learning techniques. The system is designed to address the complexities associated with identifying various types of fabrics based on their visual and structural characteristics. By integrating advanced image processing techniques, including high-resolution image acquisition, robust pre-processing pipelines, and detailed feature extraction—the system captures the critical information embedded in fabric textures, colors, and weave patterns. These extracted features are then analyzed and classified using carefully chosen algorithms such as neural networks or clustering models, optimized to handle both simple and intricate fabric types.   One of the key strengths of the system lies in its adaptability and reliability across diverse imaging conditions. The controlled acquisition setup, combined with normalization and contrast enhancement methods, ensures consistent input quality regardless of lighting variations or environmental disturbances. Furthermore, the use of OpenCV2 provides a flexible and efficient platform for image analysis, enabling real-time processing capabilities and scalability for larger datasets or commercial applications.   Through continuous testing and optimization, the system demonstrates high classification accuracy and robustness, making it a viable tool for industries such as textiles, fashion, quality control, and e-commerce. The incorporation of feedback loops and evaluation metrics like precision, recall, and F1-score allows for ongoing improvement and refinement of the model. Ultimately, this project showcases how a fusion of computer vision and machine learning can offer a practical, intelligent solution for fabric identification—simplifying manual inspection processes, reducing error rates, and accelerating fabric analysis in both research and industrial settings.

[1] Sy, H. D., Thi, P. D., Gia, H. V., & An, K. L. N. (2025). Automated fabric defect classification in textile manufacturing using advanced optical and deep learning techniques. Utilized YOLOv8, YOLOv11, RefineDet, DETR, and Faster R-CNN with optimized optical setup.   [2] Lee, S., Han, Y., & Yun, C. (2024). Development of a fabric classification system using drapability and tactile characteristics.   [3] Liu, R., Yu, Z., Fan, Q., Sun, Q., & Jiang, Z. (2024). The improved method in fabric image classification using convolutional neural network.   [4] Maray, M., Aldeib, G., Alzahrani, A., Alotaibi, F., Alsafari, S., Abdullah, E., Alghamdi, A., & Hamza, M. A. (2024). Deer hunting optimization with deep learning-driven automated fabric defect detection and classification. Used MobileNetV3, DHO for tuning, and BiLSTM for classification.   [5] T. Meeradevi , S. Sasikala (2024). Automatic fabric defect detection in textile images using a labview based multiclass classification approach. Employed object detection, classification, and preprocessing techniques for robust defect identification.   [6] Hussain, M. A. I., Khan, B., Wang, Z., & Ding, S. (2020). Woven fabric pattern recognition and classification based on deep convolutional neural networks.   [7] Oni, D. I., Ojo, J., Alabi, S., & Adebayo, A. (2018). Patterned fabric defect detection and classification (FDDC) techniques: A review.   [8] Kuo, C. F. J., Shih, C. Y., Ho, C. E., & Peng, K. C. (2010). Application of computer vision in the automatic identification and classification of woven fabric weave patterns.