Segmentation of Drilled Holes in Textured Wood Panels using Deep Learning Framework
- Publisher:
- IEEE
- Publication Type:
- Conference Proceeding
- Citation:
- 2022 IEEE 9th International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA), 2022, 00
- Issue Date:
- 2022-01-01
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| Segmentation_of_Drilled_Holes_in_Textured_Wood_Panels_using_Deep_Learning_Framework.pdf | Published version | 1.8 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
Automated vision-based detection and inspection systems become a key process in industry 4.0. Automatic precise detection of the drilled holes in various textured wood panels is an essential component in the automatic assembly lines to ensure the success of the assembly process and for the detection of the drilling defects. The wood panels are produced with a wide variety of complex textures and colors. This brings a big challenge to the vision detection process of the drilled hole panels. In this paper, we apply the deep learning mask region-based convolution neural network (Mask R-CNN) framework for the detection and segmentation of the drilled holes in synthetic images of wood panels with various textures, colors, and drilled hole patterns. A synthetic image dataset is generated for the drilled wood panels. Mask R-CNN provides simultaneous instance segmentation and detection for the drilled holes in a flexible and fast manner. The results show that the trained model can accurately perform the detection and segmentation process on the drilled holes for most images in the test dataset. We evaluate the model performance in terms of the mean average precision, the inference time, and the ability to extract the instance segmentation masks of the drilled holes.
Please use this identifier to cite or link to this item: