24 June 2024 in IEEE Transactions on Instrumentation and Measurement
In this paper, a method using a semantic segmentation method based on edge-region focal loss was proposed to estimate the heading angle of resin in a catheter-extrusion process. The approach leveraged an improved semantic segmentation facilitated by this new loss function and principal component analysis. Accurate heading angle estimation was critical and depended on the precision of segmentation, demanding robust and precise segmentation even in the presence of external disturbances. The edge-region focal loss enhanceed segmentation by heavily weighting areas with ambiguous boundaries, which was particularly important in scenarios with various semantic elements in the background and foreground or near object boundaries. Image data were collected using RGB cameras to validate the effectiveness of this method. The method’s accuracy was affirmed by the mean intersection over union (mIoU) and mean absolute error measurements, achieving mean absolute errors of the angle and mIoU at 0.5002 and 0.8657, respectively. These results demonstrate the method’s suitability for monitoring the extrusion process. Furthermore, compared to traditional loss functions, the edge-region focal loss shows superior performance in segmenting adjacent boundary regions between the background and objects and maintains robustness in noisy environments.
29 August 2023 in Advanced Materials Technologies
28 July 2025 in IEEE Access
This paper presents a real-time visual servoing system for automating die-centering in catheter extrusion processes. By leveraging state-of-the-art image processing algorithms, the system dynamically adjusts the positions of the extrusion head and tip in real time, ensuring precise alignment and significantly enhancing product consistency and quality. The proposed algorithm, a cornerstone of intelligent manufacturing, has been rigorously tested under diverse extrusion head types and operational conditions, demonstrating robust performance and precise detection of the die’s inner circle and the tip’s center position. In particular, the developed algorithm achieved a mean IoU (Intersection over Union) of 0.95, demonstrating exceptional robustness in inference performance even under dynamic conditions, such as the positional movements of the add-on system. The control signals generated by the algorithm were employed to align the tip and die, achieving precise operations at a scale of tens of micrometers, thereby validating the system’s high accuracy and reliability. Designed for seamless integration into existing manufacturing environments, the system requires minimal modifications, making it highly adaptable. By addressing long-standing challenges in die-centering operations, this intelligent manufacturing innovation not only increases production efficiency but also elevates product reliability.