I have byte[] yuvByteArray
(540x360 image captured from Camera.PreviewCallback.onPreviewFrame
method and dumped into assets/yuv.bin
fi
Kotlin
val rs = RenderScript.create(CONTEXT_HERE)
val yuvToRgbIntrinsic = ScriptIntrinsicYuvToRGB.create(rs, Element.U8_4(rs))
val yuvType = Type.Builder(rs, Element.U8(rs)).setX(byteArray.size)
val inData = Allocation.createTyped(rs, yuvType.create(), Allocation.USAGE_SCRIPT)
val rgbaType = Type.Builder(rs, Element.RGBA_8888(rs)).setX(width).setY(height)
val outData = Allocation.createTyped(rs, rgbaType.create(), Allocation.USAGE_SCRIPT)
inData.copyFrom(byteArray)
yuvToRgbIntrinsic.setInput(inData)
yuvToRgbIntrinsic.forEach(outData)
val bitmap = Bitmap.createBitmap(width, height, Bitmap.Config.ARGB_8888)
outData.copyTo(bitmap)
Our internal test app uses the following sequence to create the YUV allocation.
tb = new Type.Builder(mRS, Element.createPixel(mRS,
Element.DataType.UNSIGNED_8, Element.DataKind.PIXEL_YUV));
tb.setX(mWidth);
tb.setY(mHeight);
tb.setYuvFormat(android.graphics.ImageFormat.NV21);
mAllocationIn = Allocation.createTyped(mRS, tb.create(), Allocation.USAGE_SCRIPT);
Then inside the callback that new YUV data is available do
mAllocationIn.copyFrom(theYuvByteArray);
yuv.bin file is definitely in NV21 format, as it captures here http://developer.android.com/reference/android/hardware/Camera.PreviewCallback.html#onPreviewFrame
setYuvFormat method is from API level 18, I removed it
So this code works fine:
rs = RenderScript.create(this);
yuvToRgbIntrinsic = ScriptIntrinsicYuvToRGB.create(rs, Element.U8_4(rs));
Type.Builder yuvType = new Type.Builder(rs, Element.U8(rs)).setX(yuvByteArray.length);
Allocation in = Allocation.createTyped(rs, yuvType.create(), Allocation.USAGE_SCRIPT);
Type.Builder rgbaType = new Type.Builder(rs, Element.RGBA_8888(rs)).setX(W).setY(H);
Allocation out = Allocation.createTyped(rs, rgbaType.create(), Allocation.USAGE_SCRIPT);
in.copyFrom(yuvByteArray);
yuvToRgbIntrinsic.setInput(in);
yuvToRgbIntrinsic.forEach(out);
Using API18+ wilddev´s answer can be slightly improved as you don´t need Type.Builder objects. Do all this stuff in the onCreate() method:
aIn = Allocation.createSized(rs, Element.U8(rs), H*W*3/2); // what the f**k ? This is 12 bit per pixel, giving the length of the camera data byte array !
bmpout = Bitmap.createBitmap(W, H, Bitmap.Config.ARGB_8888); // you will need this bitmap for output anyway
aOut = Allocation.createFromBitmap(rs, bmpout); // tada !
// and set the script´s input
yuvToRgbIntrinsic.setInput(aIn);
// as pixel data are stored as int (one byte for red, green, blue, alpha), you first need an int[] array
int rgba[] = new int[w*h]; // the rgba[] array
Now you can continue with these lines or put them into the onPreviewFrame() method:
aIn.copyFrom(data); // or aIn.copyFromUnchecked(data); // which is faster and safe for camera data
yuvToRgbIntrinsic.forEach(aOut); // execute the script
aOut.copyTo(bmpout); // copy result from aOut to bmpout
// if you need the rgba-values, do
bmpout.getPixels(rgba, 0, W, 0, 0, W, H);
// now you may loop through the rgba[] array and extraxt the r,g,b,a values
// and put them into a byte[] array(s), BUT this will surely have impact on the performance when doing in Java
Change your output type passed to the constructor of ScriptIntrinsicYubToRGB
to be Element.U8_4
instead of Element.RGBA_8888
. You'll need the same for the Type.Builder
used to create the output Allocation
.