[examples] Add example programs for AprilTags detection

[fovea1959]

Added C++ and Java examples for AprilTags detection. Tested in simulator, but not on roboRIO (yet).

[Starlight220]

• Remove the "complete list" tag from the example JSONs. It's meaningless. Perhaps add a "April Tags" example instead.

• Perhaps rename the example "AprilTagVision" for consistency with the "SimpleVision" and "IntermediateVision" examples.

• The description feels rather wordy; I feel like it can be shortened. I think the key phrase is "on-RIO AprilTag detection", which is missing.

• We might also want to mention in the example's documentation that it isn't very efficient.

[fovea1959]

• Remove the "complete list" tag from the example JSONs. It's meaningless. Perhaps add a "April Tags" example instead.

both done.

• Perhaps rename the example "AprilTagVision" for consistency with the "SimpleVision" and "IntermediateVision" examples.

It's not just vision, there is more.

• The description feels rather wordy; I feel like it can be shortened. I think the key phrase is "on-RIO AprilTag detection", which is missing.

Done.

• We might also want to mention in the example's documentation that it isn't very efficient.

Is there a more efficient alternative to point someone too?

[Starlight220]

It's not just vision, there is more.

What more?

Is there a more efficient alternative to point someone too?

Using a coprocessor, same as always. The RIO is notoriously bad at vision processing.

[fovea1959]

It's not just vision, there is more.

What more?

...some analysis of the acquired images. All the other *Vision just acquire the image, perhaps statically mark it up, and send it to the dashboard.

Is there a more efficient alternative to point someone too?

Using a coprocessor, same as always. The RIO is notoriously bad at vision processing.

I would argue the RIO is slow, not inefficient; less hardware/complexity = less performance, but the ratio of complexity to performance is not bad.

...but in the interests of getting this example in the hands of teams that are already asking for it and do not have the resources to do a coprocessor (where the heck can you get a rPi right now?), I'll add something. Where do you want that warning, in the source, or in the description that we are trying to keep not-wordy? Do we also need to add it to the other vision examples?

[fovea1959]

How do we find out? I have no Mac... Can we ship it as is, and figure out the Mac issue later? I hate to delay getting something useful into the existing Rio/Linux/windows user base so that we can add vision capability (vision) to the (small) Mac user base. Alternatively, can we find a Mac tester?

[fovea1959]

• Perhaps rename the example "AprilTagVision" for consistency with the "SimpleVision" and "IntermediateVision" examples.

done.