| Autor: |
Abdullah, Raiyaan, Ahmed, Riad, Jamal, Lafifa |
| Zdroj: |
SN Computer Science; November 2022, Vol. 3 Issue: 6 |
| Abstrakt: |
This paper introduces an IoT based medicine consumption which enables the elderly person to take the right medicine at the right time. Usually elder people need to take medicine several times daily. And there is a high chance to forget the medication time. Sometimes they also struggle to remember the name of the medicine. Our system tackles this problem and ensures proper monitoring of medicine pill consumption. The system provides convenient hardware and software support to serve the medicine in a user friendly way. It contains an effective image preprocessing pipeline along with Optical Character Recognition (OCR) to detect names from medicine strips. This is also robust in detecting the medicine even if part of the strip is torn. To match the detected text with the prescribed medicine in the database Bi-directional Long Short-Term Memory (BiLSTM) networks are used. The proposed system is lighting invariant as users might consume medicine at different times of the day. Testing was performed in three different conditions—normal, medium and low lighting. The preprocessing technique plays a crucial role to detect the medicine in every lighting condition. The system also has an easy to use interface which is built using modern web technologies to enter or update the medicine consumption time. The proposed system is better than existing solutions, as it contains an array of different features including medicine dispensation, storing records, medicine detection, medicine consumption history, etc. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
Full text from SpringerLink