Apps Plants Detection

[JuliaZ]

Hello everyone!
I know, that among you there are great app developers. So, I would like to create simple Android app (version 6+) to detect plant diseases. It would be capable to use both photoes and saved images in phone memory. Model now is saving in .tflite (.lite) format. It was pre-trained on 30 classes. All neccessary model files would be done.

If it is interesting for you, please inform me!

 

[Deleted User]

I'm assuming you're asking for someone to implement this app for you?

If you want help in developing it yourself, it would be helpful if you could ask some more specific questions.

 

[JuliaZ]

Now I am trying to integrate my model to simple preparied Android app...

 

[Deleted User]

If you get stuck on a specific problem, please include any relevant code. Use [code] [/code] tags to format it.

Thanks.

 

[JuliaZ]

Ok!

 

[Deleted User]

You know there's some good documentation for TensorFlow, if that is what you're using. Including a complete sample application that would be good to get you started

https://www.tensorflow.org/lite/guide/android

 

[JuliaZ]

https://codelabs.developers.google.com/codelabs/tensorflow-for-poets-2-tflite/#4
https://github.com/amitshekhariitbhu/Android-TensorFlow-Lite-Example

I am looking throught these examples

 

[JuliaZ]

My model was trained on 30 classes, input image has a dimension of 256/256 pixels.
If it is neccessary to know other model parameters, I will write them...
There are model output labels in attached file.

 

[JuliaZ]

Also I try to intergrate my model there:
https://github.com/googlecodelabs/tensorflow-for-poets-2/tree/master/android/tflite
In file ImageClassifier.java I have changed this parameters:

static final int DIM_IMG_SIZE_X = 256;
static final int DIM_IMG_SIZE_Y = 256;

By the way there is a problem with Interpreter:
"interpreter(java.nio.mappedbytebuffer)' is deprecated"

ImageClassifier(Activity activity) throws IOException {
  tflite = new Interpreter(loadModelFile(activity));
  labelList = loadLabelList(activity);
  imgData =
      ByteBuffer.allocateDirect(
          4 * DIM_BATCH_SIZE * DIM_IMG_SIZE_X * DIM_IMG_SIZE_Y * DIM_PIXEL_SIZE);
  imgData.order(ByteOrder.nativeOrder());
  labelProbArray = new float[1][labelList.size()];
  filterLabelProbArray = new float[FILTER_STAGES][labelList.size()];
  Log.d(TAG, "Created a Tensorflow Lite Image Classifier.");
}

I am getting a mistake in app: "uninitialized classifier or invalid context".
What parameters do I need to change too?