Google is committed to advancing racial equity for Black communities. See how.

Transform output

The output of a CameraX use case is twofold: the buffer and the transformation info. The buffer is a byte array and the transformation info is how the buffer should be cropped and rotated before being shown to end users. How to apply the transformation depends on the format of the buffer.

ImageCapture

For the ImageCapture use case, the crop rect buffer is applied before saving to disk and the rotation is saved in the Exif data. There is no additional action needed from the app.

Preview

For the Preview use case, you can get the transformation information by calling SurfaceRequest.setTransformationInfoListener(). Every time the transformation is updated, the caller receives a new SurfaceRequest.TransformationInfo object.

How to apply the transformation information depends on the source of the Surface, and is usually non-trivial. If the goal is to simply display the preview, use PreviewView. PreviewView is a custom view that automatically handles transformation. For advanced uses, when you need to edit the preview stream, such as with OpenGL, look at the code sample in the CameraX core test app.

Transform coordinates

Another common task is to work with the coordinates instead of the buffer, such as drawing a box around the detected face in preview. In cases such as this, you need to transform the coordinates of the detected face from image analysis to preview.

The following code snippet creates a matrix that maps from image analysis coordinates to PreviewView coordinates. To transform the (x, y) coordinates with a 'Matrix', see Matrix.mapPoints().

Kotlin

fun getCorrectionMatrix(imageProxy: ImageProxy, previewView: PreviewView) : Matrix {
   val cropRect = imageProxy.cropRect
   val rotationDegrees = imageProxy.imageInfo.rotationDegrees
   val matrix = Matrix()

   // A float array of the source vertices (crop rect) in clockwise order.
   val source = floatArrayOf(
       cropRect.left.toFloat(),
       cropRect.top.toFloat(),
       cropRect.right.toFloat(),
       cropRect.top.toFloat(),
       cropRect.right.toFloat(),
       cropRect.bottom.toFloat(),
       cropRect.left.toFloat(),
       cropRect.bottom.toFloat()
   )

   // A float array of the destination vertices in clockwise order.
   val destination = floatArrayOf(
       0f,
       0f,
       previewView.width.toFloat(),
       0f,
       previewView.width.toFloat(),
       previewView.height.toFloat(),
       0f,
       previewView.height.toFloat()
   )

   // The destination vertexes need to be shifted based on rotation degrees. The
   // rotation degree represents the clockwise rotation needed to correct the image.

   // Each vertex is represented by 2 float numbers in the vertices array.
   val vertexSize = 2
   // The destination needs to be shifted 1 vertex for every 90° rotation.
   val shiftOffset = rotationDegrees / 90 * vertexSize;
   val tempArray = destination.clone()
   for (toIndex in source.indices) {
       val fromIndex = (toIndex + shiftOffset) % source.size
       destination[toIndex] = tempArray[fromIndex]
   }
   matrix.setPolyToPoly(source, 0, destination, 0, 4)
   return matrix
}

Java

Matrix getMappingMatrix(ImageProxy imageProxy, PreviewView previewView) {
   Rect cropRect = imageProxy.getCropRect();
   int rotationDegrees = imageProxy.getImageInfo().getRotationDegrees();
   Matrix matrix = new Matrix();

   // A float array of the source vertices (crop rect) in clockwise order.
   float[] source = {
       cropRect.left,
       cropRect.top,
       cropRect.right,
       cropRect.top,
       cropRect.right,
       cropRect.bottom,
       cropRect.left,
       cropRect.bottom
   };

   // A float array of the destination vertices in clockwise order.
   float[] destination = {
       0f,
       0f,
       previewView.getWidth(),
       0f,
       previewView.getWidth(),
       previewView.getHeight(),
       0f,
       previewView.getHeight()
   };

   // The destination vertexes need to be shifted based on rotation degrees.
   // The rotation degree represents the clockwise rotation needed to correct
   // the image.

   // Each vertex is represented by 2 float numbers in the vertices array.
   int vertexSize = 2;
   // The destination needs to be shifted 1 vertex for every 90° rotation.
   int shiftOffset = rotationDegrees / 90 * vertexSize;
   float[] tempArray = destination.clone();
   for (int toIndex = 0; toIndex < source.length; toIndex++) {
       int fromIndex = (toIndex + shiftOffset) % source.length;
       destination[toIndex] = tempArray[fromIndex];
   }
   matrix.setPolyToPoly(source, 0, destination, 0, 4);
   return matrix;
}