I completed a first test flight for roll stabilization for vertical mapping. It dramatically helps aiming the camera down, and should help with optimizing/maximizing line spacing and acheiving better results for any stitching and photogrammetry processing by getting generally vertical rather than slightly oblique photos.
I used microrax.com for building the mount itself, this is really handy stuff for these types of projects, it is the tiny version of T slot tubing. Once you use it for fabrication of things like these you will not want to use anything else. With the microrax, I can go in with an allen wrench and make easy adjustments to the angles and mount level even installed inside the plane.

I used Alpo's code slightly modified since I could not go full 90 degree stabilization due to the size of the airframe hole and the camera lens. (http://diydrones.com/profiles/blogs/easy-camera-roll-stabilizer-to - thanks for this!)
APM_RC.OutputCh(CH_6, constrain(g.rc_6.radio_in - (dcm.roll_sensor * -0.1),1300,1700));

The flight was not ideal for the test, most of it was turning without much straight and level, but still useful. Their are some jitters and it is by no means smooth, but it does point the camera much more straight down. The hole needs to be widened for vignetting, but this will wait till a final camera is used. Not nearly as nice as the Pteryx design with a rolling section of fuselage for stabilization, but great for a $99 foam airframe.

One of the side reasons for the gel vibration damping was to eliminate any servo noise from the mount stabilization itself, since the vibration damping occurs between the camera and the servo stabilized mount.

What techniques do others use to stabilize for vertical mapping? And if you have suggestions for improving it, please pass them along.

I am new to this site within about the last month, trying to get up to speed for aerial mapping using a UAV, currently using ArdupilotMega with a skywalker for vertical mapping. In mounting a small camera inside the fuselage of the Skywalker, I experimented with silicon gel for vibration isolation, which appears to be promising. The current testing used a Canon SD1200 IS camera because it was available, but I would plan on using something with a larger imaging sensor in the future (http://www.rcgroups.com/forums/showthread.php?t=1351104).

The test flight in the video above used the SD1200 in video mode in order to collect a continuous record of vibration and performance, stabilization was turned off for the flight.

I am now starting on roll stabilization using Alpo's code technique. I also need to solve camera triggering, hoping that an IR trigger solution will work. Since I may move to a Sony Nex-5, I need a non CHDK canon option for triggering.
The silicon gel approach looks very effective so far, in the future I will order some gel tape with adhesive backing for a more integrated solution and to adhere the camera to the mount, somewhat like double sided foam tape, except gel. The gel is very effective for high frequency vibration if the surface area used is matched well with the weight of the object. Their may be issues with lower frequency vibration, but that is likely the case with almost any design. Any suggestions, advice, and observations are appreciated.