Real-Time Edge Detection on Android Using OpenCV English | 繁體中文

This project demonstrates real-time edge detection on Android using the camera and OpenCV 4.12 SDK. The app has a clean, minimal design, making it an ideal starting point for learning how to integrate OpenCV into Android applications.

Features

• Real-time Edge Detection: Processes live camera frames using the Canny algorithm.
• Full-Screen Display: Displays only the camera view in a minimal layout.
• Auto-Rotation: Automatically rotates the screen based on device orientation (fullSensor).
• Modular Image Processing: Organizes image algorithms for easy extension.
• Ultra-Simple Design: Maintains a minimal and clear project structure and code.

App Screenshots

Camera preview

Canny edge detection

Core Design

• OpenCV for Android: Open-source library for image processing.
• MainActivity.java Structure:
  • cannyedge-camerax — Uses the CameraX lifecycle to manage the camera and processes images with ImageAnalysis.
  • cannyedge — Extends AppCompatActivity and follows the Android Activity lifecycle. For extending CameraActivity, refer to MainExtendsCam.java.
  • Image algorithms are organized in a modular way.
• Minimum Android Version: Android 7.0 (API level 24).

Getting Started

Follow the steps below to clone and run this project on your own system.

Development Environment

• Android Studio Narwhal 3
• OpenJDK 17 — OpenCV Prerequisite
• opencv-4.12.0-android-sdk — Download File

Project Structure

YourProject/
├── assets/
├── cannyedge/
├── opencv-sdk/     <-- add module dependency
├── README.md
└── settings.gradle

Importing and Building the Project

Clone the project

1. Import this project as a module in Android Studio.

   • Click File -> New -> Import module... and select cannyedge path

2. Follow the OpenCV official tutorial to import the OpenCV SDK as a module.

   • Click File -> New -> Import module... and select OpenCV SDK path

3. Ensure that settings.gradle correctly includes the modules:

   include ':opencv-sdk'
   include ':cannyedge'

4. Follow the OpenCV official tutorial to add the OpenCV module to this project as a Module dependency.

   • Click File -> Project structure... -> Dependencies -> All modules -> + (Add Dependency button) -> Module dependency

5. Ensure that build.gradle properly sets the OpenCV SDK as a module dependency:

   dependencies {
       implementation project(':opencv-sdk')
   }

   CameraX project only: add dependencies (Click to expand)

   dependencies {
       def camerax_version = "1.3.1"
       implementation "androidx.camera:camera-core:${camerax_version}"
       implementation "androidx.camera:camera-camera2:${camerax_version}"
       implementation "androidx.camera:camera-lifecycle:${camerax_version}"
       implementation "androidx.camera:camera-view:${camerax_version}"
   }

6. Follow the OpenCV sample script, and make sure the build.gradle includes the following configuration:

   buildTypes {
       debug {
           packagingOptions {
               doNotStrip '**/*.so'  // controlled by OpenCV CMake scripts
           }
       }
       release {
           packagingOptions {
               doNotStrip '**/*.so'  // controlled by OpenCV CMake scripts
           }
       }
   }

7. Build and run — the app will display the real-time Canny edge detection output on your device.

Application Black Screen Issues

Cause	Solution
OpenCV not initialized	Call OpenCVLoader.initLocal() in onResume()
Camera permission not granted	Request permission using requestPermissions()
setCameraPermissionGranted() not called	Add it before enableView()

Acknowledgements

This project was developed by AdevLog. Documentation and code suggestions were provided by ChatGPT, Google Gemini, and Claude.

License

This project is licensed under the MIT License. See the LICENSE file for details.