[quote author=Fu-Kwun Hwang link=topic=2231.msg8365#msg8365 date=1311295603]
I can not image there is a spatial distribution of photon on the screen from the experimental you described:

There are two slits: A and B. Photon could select  to pass either slit A or slit B.
The number being recorded are (Possible results are)
1. photon pass slit A and detector A find it.
2. photon pass slit A and detector A did find it.
3. photon pass slit B and detector B find it.
4. photon pass slit B and detector B did find it.
 
I can not find the connection between what you have described and results shown in the attached picture.
[/quote]

There has always only been one detector, located between the shutters.

When you say "2. photon pass slit A and detector A did find it." do you actually mean "2. photon pass slit A and detector A [b]didn't[/b] find it. This is completely equivalent to the point on the screen being yellow when shutter two is opened, and the detector never found it, indicating the atom went passed through slit A (the top slit), and not slit B (the bottom slit).

What is different between what the scan says, and what I've drawn on my image, is that for when what I've just described occurs (non-detection of a photon when one shutter is opened), we are colour coding those, instead of the usual X to indicate whether it goes through slit A (by detecting it from the top cavity, or not detecting it in the bottom cavity), and vice versa for the O.

When you do erasure of information - when both slits are opened the detector can click (to indicate constructive interference: remember, the photon does not exist in either cavity until you find it there, so when you open both shutters, you are getting wave interference from the possibility of the photon being in either cavity), or no click, which would indicate destructive interference -- each with 50% probability occuring. 50% of the time constructive interference occurs, 50% of the time destructive interference occurs. [see 4--> of the image]

Hope I've clarified things a bit further. :)