A deep learning framework based on Mask R-CNN (mask region-based convolutional neural network) was applied to detect and segment drilled holes in wood panels, addressing a key inspection challenge in automated assembly lines within industry 4.0 manufacturing. Wood panels are produced with highly varied textures, colors, and hole patterns, making reliable vision-based detection difficult for conventional methods. A synthetic image dataset of drilled wood panels was generated to train and evaluate the model, with Mask R-CNN enabling simultaneous, fast instance segmentation and detection of the holes supporting automated quality control and defect detection in the drilling and assembly process.
Muneer M. Al-Zu’bi, Peter Plapper, “Segmentation of Drilled Holes in Textured Wood Panels using Deep Learning Framework”, The IEEE International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA 2022), Chemnitz, Germany, 2022.
