Abstract:Visual surface defect inspection provides an important tool for product quality assessment in a wide range of industrial applications. In recent years, diverse convolutional neural network (CNN) models have been devel-oped for the high-accuracy image classification of natural objects. However, it remains a challenging task to develop deep-learning-based approaches for surface defect inspection of industrial products. The main challenge is that surface defects of industrial products occur in low probability, therefore it is impractical to build a large image dataset containing different types of defects to train/fine-tune deep CNN models to achieve satisfactory performance when applied to industrial inspection tasks. To overcome the above-mentioned problem, we build a light-weight defect classification model based on the pre-trained SqueezeNet architecture and present three effective techniques to achieve high-accuracy defect classification using very few collected defective training samples. First, we experimentally demonstrate that it is reasonable to freeze shallower convolutional layers and only fine-tune deeper layers in the last convolutional stage when there are not enough training samples. Second, we integrate batch-size independent Group Normalization (GN) into SqueezeNet to compute stable statistics based on a limited number of defective samples for training the classification model. Third, we add an auxiliary task to perform the similarity comparison of two defective samples during the training phase, significantly increasing the size of the training data and generating more distinctive defect-specific representations. The experimental results on the NEU-CLS dataset and USB-FS dataset verify the effectiveness of the proposed method. Our method can obtain accurate defect classification results when very few defective training samples are provided, i.e., about 97.69% and 82.92% average classification accuracy on the NEU-CLS and USB-FS datasets, respectively, under the 5-shot setting (5 images per defect category).

Detection Of Micro-Defects On Metal Screw Surfaces Based On Deep Convolutional Neural Networks

Detection of Defects on Metal Surfaces Based on PoolFormer Auto-Encoder

High-Accuracy Detection and Classification of Defect and Deformation of Metal Screw Head Achieved by Convolutional Neural Networks

A deep-learning-based approach for fast and robust steel surface defects classification

A Deep Learning-Based Surface Defect Inspection System Using Multiscale and Channel-Compressed Features

Automated Metal Surface Flaws Detection Using Convolutional Neural Network and Deep Visualization Analysis

An Effective Industrial Defect Classification Method under the Few-Shot Setting Via Two-Stream Training

Customized Convolutional Neural Networks Technology for Machined Product Inspection

Visual Defect Inspection of Metal Part Surface via Deformable Convolution and Concatenate Feature Pyramid Neural Networks

Metal Surface Defect Detection Method Based on Improved Cascade R-CNN

A Deep Learning-Based Approach for the Automated Surface Inspection of Copper Clad Laminate Images

A Light-Weight Deep-Learning Model with Multi-Scale Features for Steel Surface Defect Classification

Simultaneous Detection of Defects in Electrical Connectors Based on Improved Convolutional Neural Network

An Efficient Network for Surface Defect Detection

A New Steel Defect Detection Algorithm Based on Deep Learning

A Lightweight Deep Learning Network with an Optimized Attention Module for Aluminum Surface Defect Detection

A Study of Defect Detection Techniques for Metallographic Images

A Defect Detection Method for Rail Surface and Fasteners Based on Deep Convolutional Neural Network

Deep Learning Applied to Defect Detection in Powder Spreading Process of Magnetic Material Additive Manufacturing

An efficient lightweight convolutional neural network for industrial surface defect detection

Defect Detection of Mechanical Design Products Using Deep Neural Network