The textile industry has a standing history and offers a diverse range of products, such as clothing, home furnishings, and more. The quality of the product is undeniably impacted by the fabric’s quality. The caliber of the merchandise is undeniably impacted by the fabric’s quality. Still, several complications can surface during the fabric production process due to factors, like imperfect yarn or mechanical glitches. Neglecting these flaws can incur losses and harm the brand’s reputation. At times, researchers have put forth machine learning techniques, like CNN, RCNN, and others to detect defects in fabrics. However, there is still a challenge in achieving time and accurate segmentation and classification of these defects. This research paper introduces an new approach that combines the power of DCNNs with the efficiency of the YOLOv7 detection mechanism to bridge this gap. We propose a network architecture that includes preprocessing feature extraction using ResNet-50 and defect detection using YOLOv7 and then classify the fabric defect into their category using a deep convolutional neural network. This results in classification of fabric defects. Impressively, this network achieves an accuracy of 96.29%. By conducting this study, we hope to raise the bar for fabric defect detection and offer the textile industry a reliable, automated, and extremely precise instrument for ensuring fabric quality.