readme.md

Java RealSense Application for Android

This sample demonstrates how to configure an Android application to add a RealSense dependency.
The sample also demonstrates basic streaming using the Pipeline class and how to respond to connection events of a RealSense device.

Read about Android support here.

Project Configuration

Before jumping to the code, let's review the required modifications both for the module build.gradle file and to the app's manifest.

Module Build Gradle

The app should add RealSense dependency in the module build.gradle which is currently available in a private repo.

repositories {
  maven{
       url "https://egiintel.jfrog.io/artifactory/librealsense"
   }
}

dependencies {
   implementation 'com.intel.realsense:librealsense:2.+@aar'
}

The example shows how to use any librealsense version which is greater than 2.0.0, it is recommended to specify a specific version for example:

   implementation 'com.intel.realsense:librealsense:2.18.0@aar'

Once the gradle changes are done and the gradle sync was executed, you should be able to import librealsense classes. It might be required to invalidate caches in case your app is failing to import librealsense (File->Invalidate Caches->Invalidate and restart)

Application Manifest

A modifications is required in the application manifest file.

Camera Permissions

This permission is required by Androis >= 9, for apps that target lower Android version this is not required.

   <uses-permission android:name="android.permission.CAMERA"/>

Example Code

Let's look at the only source code file in this example, the MainActivity.

Note: Since the SDK is holding-on to native hardware resources, it is critical to make sure you deterministically call close on objects, especially those derived from Frame. Without releasing resources explicitly the Garbage Collector will not keep-up with new frames being allocated. Take advantage of try(){} whenever you work with frames.

On Create

RsContext.init must be called once in the application's lifetime before any interaction with physical RealSense devices. For multi activities applications use the application context instead of the activity context.

   RsContext.init(getApplicationContext());

Register to notifications regarding RealSense devices attach/detach events via the DeviceListener. In this example we start the streaming thread once a device is detected.

   mRsContext = new RsContext();
   mRsContext.setDevicesChangedCallback(new DeviceListener() {
       @Override
       public void onDeviceAttach() {
           mStreamingThread.start();
       }

      @Override
      public void onDeviceDetach() {
          mStreamingThread.interrupt();
      }
   });

Streaming

Once a device was detected we start streaming using a dedicated thread.

Streaming Thread

   private Thread mStreamingThread = new Thread(new Runnable() {
       @Override
       public void run() {
           try {
               stream();
           } catch (Exception e) {
               e.printStackTrace();
           }
       }
   });

Streaming Function

Start streaming using a Pipeline, wait for new frames to arrive on each iteration and display the distance to pixel in the center of the depth frame

   //Start streaming and print the distance of the center pixel in the depth frame.
   private void stream() throws Exception {
       Pipeline pipe = new Pipeline();
       pipe.start();
       final DecimalFormat df = new DecimalFormat("#.##");

       while (!mStreamingThread.isInterrupted())
       {
           try (FrameSet frames = pipe.waitForFrames()) {
               try (Frame f = frames.first(StreamType.DEPTH))
               {
                   DepthFrame depth = f.as(Extension.DEPTH_FRAME);
                   final float deptValue = depth.getDistance(depth.getWidth()/2, depth.getHeight()/2);
                   runOnUiThread(new Runnable() {
                       @Override
                       public void run() {
                           TextView textView = findViewById(R.id.distanceTextView);
                           textView.setText("Distance: " + df.format(deptValue));
                       }
                   });
               }
            }
        }
        pipe.stop();
   }