Research on Machine Vision-Based Detection Method for Status Monitoring of Printhead in Inkjet 3D Printing

This dataset supports the research titled "Research on Machine Vision-Based Detection Method for Status Monitoring of Printhead in Inkjet 3D Printing". It includes: (1) raw and preprocessed jetting images captured from transparent inkjet nozzles; (2) output results from the Pixel-weighting and Multi-feature Fusion (PMF) algorithm used for initial rule-based defect evaluation; and (3) training and validation datasets for lightweight convolutional neural networks (CNNs), along with corresponding classification labels and model outputs. These data were used for nozzle-level defect detection and printhead health assessment, enabling accurate classification of normal, intermittent, and failed jetting states. The dataset facilitates further research in intelligent fault diagnosis, image-based monitoring, and machine learning applications in inkjet-based additive manufacturing.