habisravi's profile - activity

I am doing a client-server image matching project for Augmented Reality. I am using Android phone as the client device. A template image is stored at the server. I have modified the OpenCV sample-Mixed Processing to retrieve the detected feature point as an array. I will send the array to the server. Is it possible to perform SURF matching with the received array and image stored at the server. If possible what are the steps to be done..

1. Try downloading ADT-Bundle from http://developer.android.com/sdk/index.html If you are using 64 bit download 64 bit version.
2. Download NDK from http://developer.android.com/tools/sdk/ndk/index.html

3. Download OpenCV for Android from http://sourceforge.net/projects/opencvlibrary/

4. Open your Home Folder in Linux /C drive in Windows.

5. Create a Folder "Android" to keep SDK, NDK, OpenCV SDK. Open the created Folder.
6. Extract ADT-Bundle to "Android" folder.
7. Extract NDK to "Android" folder.
8. Don't rename the extracted folder. If you rename,You need to change Eclipse Prefernce. Open Eclipse>Window>Preferences>Android.

9. In the Android Tab check your SDK location. If not correct Browse to extracted SDK location.

10. To Add NDK Location open Eclipse>Window>Preferences>Android>NDK. Browse to extracted NDK folder.

11. Extract OpenCVAndroid to "Android" folder.
12. Now Open Eclipse>FIle>Import>Existing Android Code into WorkSpace> Browse to "Android/OpenCV-2.4.8-android-sdk/samples/". Select All project, Tick the CheckBox"Copy Project to WorkSpace". 13.Most probably you might get a lot of errors. 14.First you need to remove errors from "OpenCV Library - 2.4.8" project.
13. Right click "OpenCV Library - 2.4.8">Android Tools>Fix Project Properties. if errors are not removed try deleting the contents of "gen" folder.
14. For all other sample projects Right click the project, click Properties>Android>Library remove the library reference and click Add>Select OpenCV Libr ary - 2.4.8 17.If Error exist delete the contents of "gen" folder. 18.Depending on the Platform set NDK path for Ubuntu http://stackoverflow.com/questions/19608610/how-to-set-android-sdk-ndk-and-jdk-paths-in-ubuntu-12-04 for Windows set NDK and OpenCV path in WindowsEnvironment.

What's your SDK Platform Windows/Linux/Mac.

1. Try downloading ADT-Bundle from http://developer.android.com/sdk/index.html If you are using 64 bit download 64 bit version.

I want to down scale the captured image to 640x320 , convert it to greyscale and save it to sdcard. Do i need to work with byte[] data for conversion. I want to use the final output image for image matching. Here is my CameraVIew class

public class CameraView extends JavaCameraView implements PictureCallback {private static final String TAG = "CameraView";
private String mPictureFileName;

public CameraView(Context context) {
    super(context, 0);
}

public void takePicture(final String fileName) {
    Log.i(TAG, "Taking picture");
    this.mPictureFileName = fileName;

      mCamera.setPreviewCallback(null);


    mCamera.takePicture(null, null, this);
}

@Override
public void onPictureTaken(byte[] data, Camera camera) {
    Log.i(TAG, "Saving a bitmap to file");

    mCamera.startPreview();
    mCamera.setPreviewCallback(this);


    try {
        FileOutputStream fos = new FileOutputStream(mPictureFileName);

        fos.write(data);
        fos.close();

    } catch (java.io.IOException e) {
        Log.e(TAG, "Exception in photoCallback", e);
    }

}

}

I want to down scale the captured image to 640x320 , convert it to greyscale and save it to sdcard. Do i need to work with byte[] data for conversion. I want to use the final output image for image matching. Here is my CameraVIew class

public class CameraView extends JavaCameraView implements PictureCallback {private static final String TAG = "CameraView";
private String mPictureFileName;

public CameraView(Context context) {
    super(context, 0);
}

public void takePicture(final String fileName) {
    Log.i(TAG, "Taking picture");
    this.mPictureFileName = fileName;

      mCamera.setPreviewCallback(null);


    mCamera.takePicture(null, null, this);
}

@Override
public void onPictureTaken(byte[] data, Camera camera) {
    Log.i(TAG, "Saving a bitmap to file");

    mCamera.startPreview();
    mCamera.setPreviewCallback(this);


    try {
        FileOutputStream fos = new FileOutputStream(mPictureFileName);

        fos.write(data);
        fos.close();

    } catch (java.io.IOException e) {
        Log.e(TAG, "Exception in photoCallback", e);
    }

}

}

I am developing an Augmented Reality Application. I am trying to use SURF for detecting objects. I have developed a custom java camera activity. I don't know how to add OpenCV SURF detector to my application. Do i need to implement SURF using NDK native development? I want to detect features in real time.

What is the best approach for processing large no of images? Storing them as Mat in database / Storing them in a folder, And retrieving them for processing.

What is the best way to store image in database for template matching? I considering to store image Mat in database.

I am doing an Augmented Reality project on Android.The user can tag images captured by a camera with a message. this message is send to the cloud. Again when the user shows the camera to the image, the message needs to be retrieved from the cloud and is overlayed above the image. Which Algorithm should i need to implement for image matching and How can i do that?

What about using pHash / perceptual Hash?

How to check whether two images are similar or not? One image is captured by a cellphone and other is stored at cloud, without sending the whole image to cloud?I need to perform this at real-time.