I'm considering a project that uses a 1080p videocam to shoot a sequence of narrow vertical images,
which will be concatenated to form a still-image 'traversorama'.

I've seen several useful tips and infobits in this forum, but have some questions before I get in deep.

I was expecting to be able to extract image strips of 1920 x <some small number> of pixels (Width) from every Nth frame (Stride), but in one post I see that the Canon HV20, which I'm considering, has a sensor that 'is optimized for' 1440x1080p resolution. If there is any interpolation happening, I'd prefer to do that when post-processing the eventual still image in Photoshop. Perhaps there is a camera that shoots native 1920x1080 - I'll research that).

So can/should I extract each strip as 1440xWidth rather than 1920x1080 or 1920xWidth?

Might it be easier to extract a sequence of 1920x1080 or 1440x1080 stills & process those later using e.g. Python Image Library (I have some rusty Python experience), or extract the narrow strips direct off the AVI(?) file?

NB I plan to shoot with the camera rotated 90 degrees, then extract horizontal strips of 'Width' lines, which will be rotated again to become vertical strips in the final traversorama.

Width, Stride and 'drive-by scanning speed' to be determined by experiment with a quick-&-dirty prototype, which will also indicate if such an approach can provide sufficient resolution in the final image

Suggestion & comments welcomed.