Fabric Defect Detection

Computer Vision & Deep Learning

Project Overview

Developed an automated visual inspection system using advanced deep learning models to enhance fabric quality control and reduce labor costs. The system accurately detects and classifies various types of fabric defects in real-time.

Key Features

YOLOV8 & SAM Integration

Advanced object detection and segmentation

TILDA Dataset Processing

Comprehensive fabric defect dataset handling

5 Defect Classes Detection

Accurate classification of multiple defect types

RoboFlow Implementation

Streamlined model training and deployment

Technology Stack

  • YOLOV8
  • Segment Anything Model (SAM)
  • Python
  • OpenCV
  • RoboFlow

Impact

  • Enhanced fabric quality control
  • Reduced labor costs
  • Accurate real-time defect detection