BoofCV works without modification on Android systems, however to build in image types are not directly compatible with BoofCV. To get around this issue BoofCV provides code in its integration package for converting to and from several Android image types. Support is currently provided for converting Bitmap and NV21 images. These functions are many times faster than trying to convert images using Bitmap's RGB values or through some more convoluted process like NV21->jpeg->Bitmap->BoofCV. Functions have been tested on Android 2.3.3 API and will most likely work on even earlier versions.
To use these additional functions you must either use the precompiled BoofCVAndroid.jar or compile it yourself from source code. Source code for Android integration can be found in the 'boofcv/integration/android' directory. An ant script is provided for compiling the jar file, but needs to be modified so that it points towards your Android API installation.
JDK 7 Issues
Some people appear to be having trouble using the jars provided on sourceforge with Android because they are compiled with JDK 7. This does not appear to be universally a problem and in some build systems they will work just fine. If you do have problems the easiest way to work around it is to recompile the source code using JDK 6.
A more detailed and working example for processing video images in Android can be found at the link below. Short code snippets are included in this section.
Converting Bitmap images
// Easiest way to convert a Bitmap into a BoofCV type ImageUInt8 image = ConvertBitmap.bitmapToGray(bitmap, (ImageUInt8)null, null); // If you are converting a sequence of images it is faster reuse a // previously declare image and buffer byte workBuffer = ConvertBitmap.declareStorage(bitmap, null); ConvertBitmap.bitmapToGray(bitmap, image, workBuffer); // Convert back into a Bitmap ConvertBitmap.grayToBitmap(image, bitmap, workBuffer); // another less efficient way bitmap = ConvertBitmap.grayToBitmap(image, Bitmap.Config.ARGB_8888); // Functions are also provided for multi-spectral images MultiSpectral<ImageFloat32> color = ConvertBitmap.bitmapToMS(bitmap, null, ImageFloat32.class, null);
Converting NV21 Images
Data from camera previews is made available in the NV21 image format.
// from NV21 to gray scale ConvertNV21.nv21ToGray(bytes,width,height,gray) // from NV21 to YUV Multi-Spectral ConvertNV21.nv21ToMsYuv(bytes,width,height,colorMS,ImageUInt8.class);
Besides converting images, VisualizeImageData is provided for visualizing
Android demonstration and benchmark applications are available for download from the google play store.
The demonstration visualizes several computer vision algorithms and provides a good estimate on what sort of runtime performance can be expected when using BoofCV on Android. The benchmark application is primarily a diagnostic app and if you submit results helps the BoofCV developers understand how will it performs on different devices. Additional information on each of these apps can be accessed using the links below. Full source is available for both applications.
- If you try to compile BoofCV in an Android you will get errors due to the swing code. The solution to this problem is to not compile it inside the Android app and instead just link to its jars.
- Compiler can't find ConvertBitmap? Make sure BoofCVAndroid.jar is included in your project.