Test common lines method with shifted images

[j-c-c]

We are currently using a Simulation with offsets set to zero for common lines testing in tests/test_orient_symmetric.py.

[garrettwrong]

If the tests aren't too long we should probably test both situations, perhaps via parameterization.

[janden]

Agreed.

[j-c-c]

Just so I remember when I tackle this issue, the trick for good performance of common lines methods for Simulations with non-zero offsets is to set shift_step to be sub-pixel (ie, < 1). This improves common line detection.

[garrettwrong]

Closed, completed.