About the project
Many patients, especially elderly people on long-term medicines for diabetes, BP or cholesterol, forget doses or take the wrong tablet. Paper prescriptions are hard to follow and family members often find out about missed doses too late. Prescription Scanner and Medicine Reminder Android project (app name RxRemind) solves this: the patient photographs the prescription, the app reads it with on-device OCR, matches every medicine with a medicine database and sets the reminders automatically. A linked caretaker is alerted when a dose is missed.
How it works
- The patient scans a printed prescription; Google ML Kit Text Recognition reads the text on the phone.
- A parser picks out the medicine name, strength, dose pattern (1-0-1, BD, TDS, HS), duration and before/after food.
- Fuzzy matching (Levenshtein similarity) corrects OCR mistakes against a master list of 78 medicines and their Indian brand names, and checks 32 drug interaction rules.
- The patient confirms the list and the app schedules exact alarms for every dose time.
- When the alarm rings, the patient can scan the medicine strip; the app checks that it is the right medicine before marking the dose taken.
- Missed doses are detected on the server and sent to the caretaker as alerts.
- A logistic regression model on the phone predicts the risk of missing doses next week and gives a tip.
Users
- Patient: scans prescriptions, gets reminders, marks doses, verifies strips, tracks stock, adherence and monthly PDF reports.
- Caretaker: links to patients with a code, sees today's doses, gets missed-dose and wrong-strip alerts, adds medicines and sends notes.
- Admin: manages patients, caretaker approval, medicine master, interaction rules, prescriptions, dose logs, notices and printable reports.
Highlights
- 3 roles, 22 Android screens, 12 admin pages, 31 PHP APIs, 13 MySQL tables
- 3 smart parts on the phone: OCR prescription reading with fuzzy medicine matching, strip verification, missed-dose risk prediction (86% accuracy)
- Full-screen dose alarm with snooze, skip and strip scan; reminders survive phone restart
- Adherence charts, dose calendar, streaks, badges, refill prediction and monthly PDF report
- Validation on every form in the app and on the server, hashed passwords, prepared statements
- Sample data and a demo prescription PDF ready for the college demo
What you get
- Complete Android Studio project (Java) and PHP admin panel with APIs
- MySQL database file with sample patients, medicines and 45 days of dose history
- Project report (Word, 55 pages) with all UML, DFD and ER diagrams
- Project abstract (2 pages), project explanation with viva questions, setup guide
- ML training scripts and dataset for the risk model
- Setup help for running on emulator or your own phone
Who can use this project
Suitable for BE / B.Tech, MCA, BCA, M.Tech and Diploma students of Computer Science, IT and related departments who need an Android project with source code, an OCR based health app, or a machine learning project with documentation and report.
Frequently asked questions
Does the OCR work without internet? Yes. ML Kit text recognition runs on the phone; Google Play services downloads the model once on first use.
Can it read handwritten prescriptions? It is built for printed or clearly written prescriptions. Every medicine can be edited before saving, so the patient always confirms the list.
Do I need a real phone for the demo? An emulator works for all screens. A real phone on the same Wi-Fi as the laptop gives the best scanning demo.
Is the ML model trained by me? The Python training script and dataset are included, so you can retrain it and show the accuracy and confusion matrix in your viva.
