About the project
Skin problems such as acne, eczema, fungal infection and psoriasis are very common, but many people delay meeting a dermatologist. Skin disease detection using CNN Android project – TvachaScan – lets a patient photograph the affected skin and get an instant screening on the phone itself. A MobileNetV2 CNN (TensorFlow Lite) classifies the photo into 9 common condition groups and shows the top-3 results with confidence. Precautions, a care plan and a link to real dermatologists then turn that result into guidance the patient can act on.
How it works
- Patient takes or picks a skin photo; the app checks blur, light and skin coverage first
- The on-device CNN predicts the condition with top-3 predictions and a confidence ring
- Affected area %, spot count and severity (Mild / Moderate / High) are calculated
- A 10-question self-assessment gives a Low / Moderate / High risk score
- Precautions and a 4-week AI care plan are shown; serious cases go to a dermatologist
- The dermatologist confirms or corrects the result and adds a prescription
- Patient books a slot, tracks healing with charts and downloads a PDF skin report
Users
- Patient: skin scan, result, care plan, self-test, history, healing tracker, booking, routine streak, PDF report
- Dermatologist: review queue, diagnosis + prescription, appointment requests, patient history, working days and slots
- Admin: dashboard, approvals, conditions, precautions, questions, scans, appointments, notices, reports, settings
Highlights
- 3 roles, 23 app screens, 15 admin pages, 31 PHP APIs, 13 database tables
- TFLite CNN (MobileNetV2 transfer learning) with Colab training notebook included
- Photo quality check, severity estimation, risk scoring and offline quick-check mode
- Healing tracker with before / after photos and trend chart, routine streaks and badges
- Validations on every form in the app and again on the server (email, 10-digit mobile, pincode, password, dates, photo)
- Passwords hashed, prepared SQL statements, role check on every API
- Sample data and demo logins so every screen shows data in the demo
What you get
- Complete Android Studio source code (Java) and PHP admin panel + APIs
- MySQL database file with sample data
- Model training notebook and script (DermNet, 9 classes)
- Project report (Word, 50+ pages) with all UML and DFD diagrams
- 2-page project abstract, project explanation with viva questions
- Setup guide (PDF + Word) and setup help
Who can use this project
Final year students of BE / B.Tech, MCA, BCA and Diploma in Computer Engineering, IT and related branches who want an Android project with source code, a machine learning (CNN) project, or a healthcare app project with documentation and report.
Frequently asked questions
Does the CNN need internet to predict? No. The TFLite model runs inside the phone; internet / Wi-Fi is needed only to save data on the server.
Which skin conditions are covered? Acne, eczema, fungal infection, psoriasis, pigmentation / vitiligo, moles (melanoma risk), scabies / bites, urticaria and viral warts.
Can I train the model again with my own images? Yes. The included Colab notebook trains MobileNetV2 and exports skin_model.tflite and labels.txt.
Will it run on a real phone? Yes. Connect the phone and laptop to the same Wi-Fi and enter the laptop IP on the login screen.
