Pill Detection Model for Medicine Inspection Based on Deep Learning
importance 4/10
ai assisted
Industrial Vision Systems and Defect Detection pharmacy automation stage: research
Problem
Detecting individual pills in images containing multiple pills is difficult when training data is limited and pill combinations grow exponentially with pill types.
Why it matters
Manual inspection of dispensed pills risks human error and productivity loss, with potential patient harm.
What the AI did
A deep learning algorithm with a database expansion method detects individual pills in multi-pill images using limited training data.
Results
not stated in source
Impact
Authors claim the method outperforms existing algorithms and will help minimize inspection errors and productivity loss.
Record
| Venue | Chemosensors |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Hyuk‐Ju Kwon, Hwi-Gang Kim, Sung-Hak Lee |
| Citations | 48 (OpenAlex, at import time) |
| References | 32 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/chemosensors10010004 |
| Open access | gold |
| Keywords | pill detection, deep learning, medicine inspection, data expansion |
Importance rating
4
real but narrower advance
Pill inspection errors risk patient harm and productivity loss in dispensing; a real but narrow advance with no stated scale of affected patients or pharmacies.