TPU Film Smart Defect Detection System
Based on industrial vision and deep learning for real-time defect detection and classification in TPU film production.
Industry Pain Points & Challenges
Traditional manual sampling only covers 5%-10%, leading to missed defects in high-speed production.
Micro defects (0.1mm) in transparent TPU film are hard for the human eye to detect stably.
Lack of digital records makes it hard to analyze defect types and optimize production processes.
Technical Architecture & Solution
Image Acquisition
Basler industrial cameras + specialized lighting capture high-definition images of TPU film.
Edge Computing
High-performance industrial PCs (i7 + RTX) process image data in real-time with low latency.
AI Inference Engine
Deep learning models (YOLO v8) accurately identify crystal points, black spots, and scratches.
Application Interaction
Real-time alerts, defect map visualization, and history reports for production decisions.