Does the camera arrive to a really close up of the picture?
If it doesn't, then I would use a virtual camera adjusting its focus to the picture, so everything else gets blurred (and the pixelation is not too blatant). I would center the camera to the frame, and use a slight transition from the picture to the footage once the image fills all the frame.
I see that the picture is held slightly backwards, so you will have to adjust the perspective in order to match the picture with the footage.
It may make the transition more seamless if you place the real photo in the frame. So you don't have to line up the real photo with the one you want to add digitally. This can be done with the 3D tracker. Check out this tutorial for more and best of luck: