Segmentation of Drilled Holes in Texture Wooden Furniture Panels Using Deep Neural Network

Abstract

Drilling operations are an essential part of furniture from MDF laminated boards required for product assembly. Faults in the process might introduce adverse effects to the furniture. Inspection of the drilling quality can be challenging due to a big variety of board surface textures, dust, or woodchips in the manufacturing process, milling cutouts, and other kinds of defects. Intelligent computer vision methods can be engaged for global contextual analysis with local information attention for automated object detection and segmentation. In this paper, we propose blind and through drilled holes segmentation on textured wooden furniture panel images using the UNet encoder-decoder modifications enhanced with residual connections, atrous spatial pyramid pooling, squeeze and excitation module, and CoordConv layers for better segmentation performance. We show that even a lightweight architecture is capable to perform on a range of complex textures and is able to distinguish the holes drilling operations’ semantical information from the rest of the furniture board and conveyor context. The proposed model configurations yield better results in more complex cases with a not significant or small bump in processing time. Experimental results demonstrate that our best-proposed solution achieves a Dice score of up to 97.89% compared to the baseline U-Net model’s Dice score of 94.50%. Statistical, visual, and computational properties of each convolutional neural network architecture are addressed.

Deep-Learning-Based Estimation of the Spatial QRS-T Angle from Reduced-Lead ECGs

“…Smaller networks like ours, or as in the one applied for automatic diagnosis of 12-lead ECGs [ 42 ], outperform their convolutional-blocks-only counterparts when enhanced with custom blocks such as residual connections, squeeze-and-excitation, atrous spatial pooling, or case-specific loss functions [ 32 , 33 ]. Our strategy involved residual blocks with a predominant focus on an original loss function.…”

“…First, complex CNNs employed for image-based applications are likely an overengineered solution for our problem. Second, smaller CNN architectures enhanced with residual connections and case-specific loss functions can outperform architectures based on regular convolutional blocks [ 32 , 33 ]. Third, lightweight and low-complexity models are preferable for deployment in devices with hardware and computational constraints, such as consumer healthcare devices.…”

A Dynamic Approach for Railway Semantic Segmentation

“…For the rails segmentation problem we have chosen models from the U-Net family, which feature an encoder-decoder architecture with skip connections between distanced layers. Although these types of architectures were designed to solve the semantic segmentation task for medical images, many studies have shown how well they work for other tasks as well [2,14]. In our experiments we have considered two model architectures: U-Net [17] and ResUNet++ [13].…”