Detection and Quantification of Visual Tablet Surface Defects by Combining Convolutional Neural Network-Based Object Detection and Deterministic Computer Vision Approaches DOI Open Access

Eric Freiermuth,

David Kohler,

Albert Hofstetter

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: 2(2), P. 9 - 9

Published: May 15, 2025

Tablet surface defects are typically controlled by visual inspection in the pharmaceutical industry. This is an insufficient response variable for knowledge-based formulation and process development, it results rather limited robustness of control strategy. In this article, we present analytical method quantitative characterization tablet defects. The involves analysis a digital microscope to obtain optical images three-dimensional scans. Pre-processing procedures applied simplification data allow detection imprint characters structures Faster R-CNN object model. Geometrical variables like perimeter area were derived from model statistically analyzed selected number tablets. allowed development product-specific acceptance criteria small reference dataset, evaluation sticking, picking, chipping, abrasion showed high precision sensitivity demonstrated robust without false negative results. image was automated, developed algorithm can be operated simple routine on standard computer few minutes. suitable industrial use enables advancements while providing novel opportunity quality

Language: Английский

A Novel Paradigm on Data and Knowledge-Driven Drug Formulation Development: Opportunities and Challenges of Machine Learning DOI
Xinrui Wang,

Zhenda Liu,

Lin Xiao

et al.

Journal of Industrial Information Integration, Journal Year: 2025, Volume and Issue: unknown, P. 100796 - 100796

Published: Feb. 1, 2025

Language: Английский

Citations

0
Detection and Quantification of Visual Tablet Surface Defects by Combining Convolutional Neural Network-Based Object Detection and Deterministic Computer Vision Approaches DOI Open Access

Eric Freiermuth,

David Kohler,

Albert Hofstetter

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: 2(2), P. 9 - 9

Published: May 15, 2025

Tablet surface defects are typically controlled by visual inspection in the pharmaceutical industry. This is an insufficient response variable for knowledge-based formulation and process development, it results rather limited robustness of control strategy. In this article, we present analytical method quantitative characterization tablet defects. The involves analysis a digital microscope to obtain optical images three-dimensional scans. Pre-processing procedures applied simplification data allow detection imprint characters structures Faster R-CNN object model. Geometrical variables like perimeter area were derived from model statistically analyzed selected number tablets. allowed development product-specific acceptance criteria small reference dataset, evaluation sticking, picking, chipping, abrasion showed high precision sensitivity demonstrated robust without false negative results. image was automated, developed algorithm can be operated simple routine on standard computer few minutes. suitable industrial use enables advancements while providing novel opportunity quality

Language: Английский

Citations

0