冻干物的拓扑结构和表面特性显著影响冻干药物的稳定性和可重构性。因此,产品的视觉质量控制势在必行。然而,此过程不仅耗时耗力，而且昂贵且容易出错。在本文中,我们提出了一种完全自动化的方法，冻干药品的无损检测，利用机器人技术，计算机断层扫描,和机器学习。
The topology and surface characteristics of lyophilisates significantly impact the stability and reconstitutability of freeze-dried pharmaceuticals. Consequently, visual quality control of the product is imperative. However, this procedure is not only time-consuming and labor-intensive but also expensive and prone to errors. In this paper, we present an approach for fully automated, non-destructive inspection of freeze-dried pharmaceuticals, leveraging robotics, computed tomography, and machine learning.
