Ok Now, Where Is It?

How can we use partial state information to find objects (particularly those with low signal-to-noise ratio - SNR)?

Timothy Murphy tells us all about it in his recently published article Space Object Detection in Images Using Matched Filter Bank and Bayesian Update.

In his paper, Tim demonstrates how to partition the unobservable state space in terms of a dissimilarity metric for the expected signature of the followup observation. He then proceeds to show how to use a Match Filter to optimally extract the expected signature from the followup electro-optical image.

Using this methodology, Tim has been able to successfully detect objects with photometric SNR < 1 and update Bayesian probability density functions after several measurements.

Congratulations Tim!