Multi-Axis 3D Printing Defect Detecting by Machine Vision with Convolutional Neural Networks

Zhang, H.; Zong, Zefang; Yao, Yuwei; Hu, Qinyong; Lammer, Herfried

doi:10.1007/s40799-022-00577-2

Cited by 11 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to traditional region-based Convolutional Neural Network (CNN) [13] and support vector machine (SVM) [14] detection methods, YOLOv8 can detect objects in real time in an image and provide their location and category information simultaneously.…”

Section: Introductionmentioning

confidence: 99%

An Improved 3D Printing Extrusion Defect Detection Method Based On YOLO-v8

Cao,

Fu,

et al. 2024

Preprint

View full text Add to dashboard Cite

Additive manufacturing (AM) technology finds widespread use in consumer and industrial manufacturing. However, large-scale 3D printers in the manufacturing industry encounter challenges like extended printing times, high nozzle flow, and maintaining flow stability. Addressing these issues necessitates thorough research and optimization of 3D printers to enhance printing efficiency and stability. This study proposes a real-time monitoring system for identifying flow defects in large 3D printers using machine vision. The system comprises a camera capturing extrusion images, a computer processing real-time screens, and a lower computer within the monitoring system designed for real-time monitoring and correction of flow quality in large 3D printers. Defects are detected and image data captured by adjusting the extrusion speed, while an attention mechanism module optimizes the YOLOv8 model. The model is trained for deep learning using mAP50 as the performance evaluation index, achieving a mean average precision (mAP) of 92.1%. Subsequently, the optimized model is deployed on a host computer. Upon detecting flow defects, it sends commands to a Raspberry Pi, generating G-code commands to adjust extrusion speed and relay them to the 3D printer. This enables real-time monitoring and automatic calibration of the printing process. The study presents an effective solution for real-time monitoring and calibration in closed-loop processes within additive manufacturing.

show abstract

Section: Introductionmentioning

confidence: 99%

An Improved 3D Printing Extrusion Defect Detection Method Based On YOLO-v8

Cao,

Fu,

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…A diverse set of image data, including various error levels and printing geometries, is required for training CNNs to ensure the system's generalization and effectiveness [19]. To make the detection process real-time, Zhang et al [20] proposed a method using machine vision and convolutional neural networks (CNNs) to detect multi-axis FDM printing defects. Their self-built CNN network achieved an 83.1% classi cation accuracy for interlayer delamination defects.…”

Section: Introductionmentioning

confidence: 99%

Real-time defect detection for FDM 3D printing using lightweight model deployment

Hu,

Chang,

et al. 2024

Preprint

View full text Add to dashboard Cite

FDM 3D printing is one of the most widely used additive manufacturing methods, bringing great convenience to production manufacturing. However, various printing defects may occur during the printing process due to human factors or printer-related issues. Timely detection of defects and halting printing becomes a scenario of significant practical importance. This paper first analyzes the causes of the five most common defects in FDM 3D printing, and a defect dataset is created by deliberately designing defects. Subsequently, a real-time defect detection system for FDM 3D printing, based on an improved YOLOv8 detection head, is developed. By employing an optimization method using Group Convolution to share parameters, the detection head is lightweight, resulting in better model performance. Experimental results demonstrate that the mAP50 of the improved YOLOv8 model reaches 97.5%, with an 18.1% increase in FPS and a 32.9% reduction in GFLOPs. This enhancement maintains comparable detection accuracy to the original model while achieving faster detection speed and lower computational requirements. The improved model is integrated into the detection system as the detection model, and through testing, the real-time detection system promptly and accurately identifies and alerts any occurring defects. The practical significance of this system lies in its ability to enhance production efficiency, reduce resource wastage due to defective printing, and improve product quality and manufacturing safety, thereby providing strong support for the application of visual inspection technology in FDM 3D printing.

show abstract