HN2new | past | comments | ask | show | jobs | submitlogin

I think it's checking how many x-rays are reflected back to the machine.

But since the background and the object are both metal they both reflect equally, and are indistinguishable.



The background is not metal, it is open space that does not reflect xrays. The targeted metal objects also do not reflect xrays.

It is more intuitive to say that these machines are not threat detectors (like metal detectors), they are people detectors. The lack of a person is considered to be a threat within the bounds of their silhouette. Place a threat outside the silhouette and it is undetectable.


So what would happen if they had a wire mesh (chicken wire, for example) for the background? An interruption in the wire mesh pattern would indicate an anomaly. They could change up the pattern every so often to thwart any attempt to match the pattern --even if it woud be difficult to match up.


The scanner images both sides of a person at the same time. To add a background, you would have to have two poses from the same side, effectively halving throughput and complicating the imaging procedure.


Can't they add some time based depth information to discriminate ?

Or add a slightly absorbing background , additional maybe. ...


Some pattern in the background maybe?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: