1Professor, Department of Information Technology, Nehru Institute of Technology, Coimbatore, TN, India.
2,3,4,5Department of Information Technology, Nehru Institute of Technology, Coimbatore, TN, India.
Email: santhan.mca@gmail.com, divyadharshini0292@gmail.com, muzzammilmohammed370@gmail.com, nandhinisridhar12@gmail.com, varshiniarunagu@gmail.com
Background: Nowadays, a large part of our population depends on daily medication, which makes the quality, authenticity, and safety of tablets be three major criteria to be met. Nevertheless, there are cases when tablets are discovered out of the blister pack, having faded imprints, looking old, or even have slightly changed their shape. Then, hardly can their legitimacy be confirmed and the only sure way becomes a laboratory test, which takes a lot of time and is not at hand immediately.
Objective: This work proposes a compact and portable pill verification device based on a dual-modal decision pipeline. The system integrates spectroscopic analysis using a micro NIR spectrometer and imaging analysis using a compact CMOS camera equipped with a macro microlens. The spectral data captures the chemical fingerprint of the tablet, while the imaging module analyzes its visual and structural characteristics.
Methodology: The collected data from both modalities are processed using a trained machine learning classifier for spectral verification and a convolutional neural network (CNN) for image-based validation. The final authentication decision is obtained through decision-level data fusion of both outputs. In addition, the proposed system can very quickly and easily identify pills in a non-destructive and portable manner, thus providing a convenient and readily available alternative to the lengthy and costly standard laboratory-based drug verification that is not easily accessible.
Conclusion: The dual-modal framework combining spectral and imaging analysis offers a rapid, non-destructive, and portable solution for tablet authentication, addressing the critical need for accessible drug verification outside laboratory settings.
Keywords: spectroscopic analysis, microlens imaging, machine learning, rapid and non-destructive tablet verification.
How to cite this article: Shanthakumar P, Divyadharshini T, Muzzammil AM, Nandhini S, Varshini AN. Pillsure: A Dual-Modal Spectral and Imaging Framework for Tablet Authentication. Int J Drug Deliv Technol. 2026;16(12s): 537-545. DOI: 10.25258/ijddt.16.12s.65
Source of support: Nil.
Conflict of interest: None