Computer Vision and Normalizing Flow-Based Defect Detection

TL;DR

This paper introduces an unsupervised, fully automated 360-degree defect detection system for manufacturing that leverages normalizing flows, object detection, and background subtraction, achieving high accuracy without altering production lines or requiring defect samples.

Contribution

The paper presents a novel three-stage plug-and-play defect detection system that operates without supervision and can be deployed on existing assembly lines with multiple cameras.

Findings

Achieves 0.90 AUROC in real-world tests

Operates without modifying assembly lines or collecting defect samples

Handles background noise, varying angles, and sizes effectively

Abstract

Visual defect detection is critical to ensure the quality of most products. However, the majority of small and medium-sized manufacturing enterprises still rely on tedious and error-prone human manual inspection. The main reasons include: 1) the existing automated visual defect detection systems require altering production assembly lines, which is time consuming and expensive 2) the existing systems require manually collecting defective samples and labeling them for a comparison-based algorithm or training a machine learning model. This introduces a heavy burden for small and medium-sized manufacturing enterprises as defects do not happen often and are difficult and time-consuming to collect. Furthermore, we cannot exhaustively collect or define all defect types as any new deviation from acceptable products are defects. In this paper, we overcome these challenges and design a three-stage plug-and-play fully automated unsupervised 360-degree defect detection system. In our system, products are freely placed on an unaltered assembly line and receive 360 degree visual inspection with multiple cameras from different angles. As such, the images collected from real-world product assembly lines contain lots of background noise. The products face different angles. The product sizes vary due to the distance to cameras. All these make defect detection much more difficult. Our system use object detection, background subtraction and unsupervised normalizing flow-based defect detection techniques to tackle these difficulties. Experiments show our system can achieve 0.90 AUROC in a real-world non-altered drinkware production assembly line.