android - Mobile Vision API - concatenate new detector object to continue frame processing -

i want use new face detection feature vision api provides along additional frame processing in application. this, need have access camera frame processed face detector, , concatenate processor using face detected data.

as see in sample, camerasource abstracts detection , camera access, , can't have access frame being processed. there examples of how camera frame in api, or, maybe, create , concatenate detector receives it? possible @ least?

thanks, lucio

yes, possible. you'd need create own subclass of detector wraps facedetector , executes frame processing code in detect method. this:

class myfacedetector extends detector<face> {   private detector<face> mdelegate;    myfacedetector(detector<face> delegate) {     mdelegate = delegate;   }    public sparsearray<face> detect(frame frame) {     // *** add custom frame processing code here     return mdelegate.detect(frame);   }    public boolean isoperational() {     return mdelegate.isoperational();   }    public boolean setfocus(int id) {     return mdelegate.setfocus(id);   } } 

you'd wrap face detector class, , pass class camera source. this:

    facedetector facedetector = new facedetector.builder(context)             .build();     myfacedetector myfacedetector = new myfacedetector(facedetector);      myfacedetector.setprocessor(/* include processor here */);      mcamerasource = new camerasource.builder(context, myfacedetector)             .build(); 

your detector called first raw frame data.

note image may not upright, if device rotated. can orientation through frame's metadata.getrotation method.

one word of caution: once detect method returns, should not access frame pixel data. since camera source recycles image buffers, contents of frame object overridden once method returns.

edit: (additional notes) avoid boilerplate code of myfacedetector using multidetector this:

multidetector multidetector = new multidetector.builder()     .add(new facedetector.builder(context)                 .build())     .add(new yourreallyowndetector())     .build(); 

also note use of facetrackerfactory in conjuction multiprocessor described there.