Abstract:In practical industrial applications, the inference speed of deep learning models directly affects the efficiency of industrial production. Therefore, the lightweight real-time detection method of surface defects is an essential task in the industrial process. We need to achieve a favorable balance between efficiency and accuracy since the rising demand for production efficiency. However, most of the existing pixel-level detection methods 1) often adopt huge computational overhead to learn rich features, resulting in slow inference speed and 2) show a performance degradation when applied to different industrial surface defect scenarios. To this end, we propose an efficient targeted design (ETD) for real-time defect detection of surface defects. It consists of two branches: (i) an efficient feature enhancement branch, with global aggregation module (GAM) and cross-scale guide module (CGM) to gradually enhance defect features, and (ii) an edge posterior branch, with verification module (VM) and scale interaction module (SIM) to implicitly guide the boundary details of defects. Specifically, while inheriting this framework, we reconsider the relationship between precision, parameters, and speed so that our model can be applied to different industrial scenarios. Extensive experimental results on four datasets indicate that ETD outperforms other leading saliency detection methods. Meanwhile, our method ETD-S achieves 347 FPS on ESDIs-SOD dataset, 254 FPS on Crack500 dataset, 227 FPS on NRSD-MN dataset and 273 FPS on DAGM dataset. Additionally, we conduct real-time analysis of ETD on an intelligent paradigm for industrial surface defect detection, further demonstrating its efficacy in practical scenarios. ETD demonstrates effective detection performance while achieving a lightweight architecture, which can be implemented using various deep learning frameworks, showcasing substantial potential for real-time surface defect detection. The source code and dataset are publicly available at https://github.com/VDT-2048/ETD.

An efficient and real-time steel surface defect detection method based on single-stage detection algorithm

A Real-Time Steel Surface Defect Detection Approach with High Accuracy

An efficient targeted design for real-time defect detection of surface defects

Rtsds:a real-time and efficient method for detecting surface defects in strip steel

A real-time anchor-free defect detector with global and local feature enhancement for surface defect detection

An End-to-End Steel Surface Defect Detection Approach via Fusing Multiple Hierarchical Features

Multi-scale sensing and multi-dimensional feature enhancement for surface defect detection of hot-rolled steel strip

Strip steel surface defect detecting method combined with a multi-layer attention mechanism network

A New Steel Defect Detection Algorithm Based on Deep Learning

A Highly Efficient and Lightweight Detection Method for Steel Surface Defect

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

Surface defect detection of hot rolled steel based on multi-scale feature fusion and attention mechanism residual block

Online Detection of Surface Defects Based on Improved YOLOV3

SPEED:Semantic Prior and Extremely Efficient Dilated Convolution Network for Real-Time Metal Surface Defects Detection

Surface defect detection of industrial components based on vision

Defect Detection in Steel Using a Hybrid Attention Network

Efficient Detection Model of Steel Strip Surface Defects Based on YOLO-V7

A steel surface defect inspection approach towards smart industrial monitoring

An Algorithm for Detecting Surface Defects in Steel Strips Based on an Improved Lightweight Network

Steel Surface Defect Detection Method Based on Improved YOLOv8

A deep learning model for steel surface defect detection