📦 Inventory Scanner — Android Mobile Application

A mobile application for scanning, recognizing, and counting inventory objects using AI (Google ML Kit) and a REST API backend.

---

📖 About the Project

Inventory Scanner is an Android application developed as a diploma project at Ust-Kamenogorsk Higher Polytechnic College (2022). The application is designed for employees of organizations (warehouses, schools, enterprises) who need to conduct asset inventory efficiently.

Instead of manually typing inventory numbers, users can photograph or scan an item's inventory tag — the AI automatically recognizes the number and searches it against a remote database, then generates an inventory report.

---

✨ Features

Feature	Description
🏫 Room Selection	Choose a classroom or department from the server database
🔍 Smart Search	Live search by inventory number within a room or across the global database
📷 AI Scanner	Recognize inventory numbers from a photo (camera or gallery)
✅ Object Marking	Mark found items as checked during the inventory session
📊 Report Generation	Generate and save an inventory report to device storage
🌐 REST API Sync	All data is synced with a remote MariaDB server via HTTP

---

🏗️ Architecture

The application uses a 2-tier client-server architecture:

┌────────────────────┐         HTTP (REST API)        ┌─────────────────────┐
│   Android Client   │  ──────────────────────────►   │    Remote Server    │
│  (Kotlin + ML Kit) │  ◄──────────────────────────   │  MariaDB + PHP API  │
└────────────────────┘                                └─────────────────────┘

• Client: Android app (Kotlin, XML layouts)
• Server: Hosted on Gohost.kz, database managed via phpMyAdmin
• Communication: REST API using HTTP methods (GET, POST, PUT, DELETE)
• AI: Google ML Kit (on-device text recognition, no cloud dependency for basic scanning)

---

🗄️ Database Schema (3NF)

Classrooms          Categories          Items
──────────────      ──────────────      ──────────────────────────
num (PK)            id (PK)             inventory_num (PK)
name                name                name
                                        category_id (FK → Categories)
                                        class_room_num (FK → Classrooms)
                                        created_at
                                        updated_at

---

📱 Screens & Navigation

SplashScreen
    │  (permission check + internet check)
    ▼
MainActivity (Home)
    │  BottomNav: [🏠 Home] [🔍 Search]
    ├── HomeFragment — classroom list from server
    │       │
    │       ▼
    │   SearchFragment — items list for selected room
    │       ├── 🔍 Live search (room → global DB)
    │       ├── 📷 Scanner → ScannerActivity
    │       └── 📊 Report → CartListActivity
    │
    └── ScannerActivity — camera / gallery + crop + ML Kit OCR

---

🛠️ Tech Stack

Layer	Technology
Language	Kotlin (primary), Java, XML
IDE	Android Studio
AI / ML	Google ML Kit — Text Recognition
Networking	REST API over HTTP/HTTPS
Database	MariaDB (remote), phpMyAdmin
Architecture	2-Tier Client-Server, MVVM (ViewModel)
Min SDK	Android 8.0 (API Level 26)

---

📋 System Requirements

Device Requirements

Parameter	Minimum
Android version	8.0 (Oreo) or higher — API Level 26
RAM	1.5 GB or more
Free storage	800 MB or more
Screen resolution	1080 × 1920 px (FHD)
CPU frequency	1.6 GHz or higher
CPU cores	2 or more
Internet	Required (for DB sync and ML Kit cloud fallback)

Permissions Required

• CAMERA — for scanning inventory tags
• READ_EXTERNAL_STORAGE / WRITE_EXTERNAL_STORAGE — for gallery access and report saving
• INTERNET — for REST API communication

---

📂 Project Structure

app/
├── manifests/
│   └── AndroidManifest.xml          # App config, permissions, components
├── java/
│   ├── MainActivity.kt              # Main entry screen
│   ├── SplashScreen.kt             # Splash + permission check
│   ├── ScannerActivity.kt          # Camera/gallery scan + ML Kit OCR
│   ├── CartListActivity.kt         # Report screen (marked items)
│   ├── HomeFragment.kt             # Classroom list fragment
│   ├── SearchFragment.kt           # Item search + scan trigger
│   ├── ConnectionDetector.kt       # Internet connectivity check
│   ├── HomeViewModel.kt            # ViewModel for HomeFragment
│   ├── ScanViewModel.kt            # ViewModel for ScannerActivity
│   ├── SearchViewModel.kt          # ViewModel for SearchFragment
│   ├── adapters/                   # RecyclerView adapters
│   ├── data/                       # REST API (Retrofit/HTTP) calls
│   ├── domain/                     # Data models (Classroom, Item, Category)
│   ├── interfaces/                 # Click listeners, UI interfaces
│   └── utils/                      # Constants and utility functions
└── res/
    ├── layout/                     # XML screens (6 screens + templates)
    ├── drawable/                   # Icons, vector graphics
    ├── font/                       # 4 custom fonts
    ├── mipmap/                     # App launcher icons
    ├── values/                     # Colors, themes, strings
    └── xml/                        # Backup rules, file paths

---

🚀 Getting Started

1. Clone the repository

git clone https://github.com/deavers/InvScan.git
cd InvScan

2. Open in Android Studio

Open the project folder in Android Studio (Flamingo or newer recommended).

3. Configure the API endpoint

In utils/Constants.kt (or equivalent), set your server base URL:

const val BASE_URL = "https://your-server.com/api/"

4. Set up the database

Import the SQL schema into your MariaDB server using phpMyAdmin:

CREATE TABLE classrooms (num VARCHAR(20) PRIMARY KEY, name VARCHAR(100));
CREATE TABLE categories (id INT PRIMARY KEY AUTO_INCREMENT, name VARCHAR(100));
CREATE TABLE items (
    inventory_num VARCHAR(50) PRIMARY KEY,
    name VARCHAR(200),
    category_id INT,
    class_room_num VARCHAR(20),
    created_at DATETIME,
    updated_at DATETIME,
    FOREIGN KEY (category_id) REFERENCES categories(id),
    FOREIGN KEY (class_room_num) REFERENCES classrooms(num)
);

5. Build & Run

# Via Android Studio: Run ▶ or
./gradlew assembleDebug

---

🧠 How ML Kit Text Recognition Works

1. User takes a photo or selects from gallery
2. The image is cropped to the region of interest
3. Google ML Kit processes the image on-device:
   • If confidence ≥ 70% → result accepted
   • If confidence < 70% → fallback to Google Cloud ML Kit
   • If still < 70% → closest match is used
4. Recognized number is searched in the room DB, then the global DB
5. Result is displayed with item info and category

Average recognition time: ~1 second (varies by device and image quality)

---

📄 Report Format

Generated reports are saved to device storage as files named:

[RoomName]_[YYYY-MM-DD].xlsx / .txt

Report contents:

• Room full name
• Total items in room
• Number of marked (checked) items
• All inventory numbers with ✅ / ❌ status per item

---

🔌 REST API Endpoints

Method	Endpoint	Description
GET	/classrooms	Get all classrooms
GET	/items?room={num}	Get items for a specific room
GET	/items/{inventory_num}	Search item globally
POST	/items	Add new item
PUT	/items/{inventory_num}	Move item to another room
DELETE	/items/{inventory_num}	Remove item

---

👤 Author

Sairakhan Shynggys Diploma Project — Ust-Kamenogorsk Higher Polytechnic College
Specialty: Computer Engineering and Software
Group: 18-VT-2 | Year: 2022

---

🔗 Related Projects

This app requires a running backend server to function.

Repository	Description
BaseInvScan	Laravel REST API backend — provides inventory items, classrooms and categories

To run the full project, clone and start BaseInvScan first, then update BASE_URL in the app's constants file before building.

📜 License

This project was developed as an educational diploma project.