Smartphones today invariably come with a camera and the ability to take pictures and movies.
Building these facilities into an app is straightforward. Apple and Android both provide all the necessary software to allow apps to be built incorporating these features.
The standard product on the templedynamic website allows users to take photos and share them.
Movies are much the same a photos, but be careful: movies need larger amounts of memory on the phone, and substantially higher bandwidth to send them around. This is an important design factor in any system for video.
Cost and technical implications: Allowing your users to take photos and share them with the app owner or other users should be standard, so low to medium cost.
Using movies will be medium to high cost.
The technology for audio and video conferencing is well established. Just about all smartphones on sale today have the necessary hardware - i.e microphone, speakers and a camera if video conferencing is needed.
To set up audio and video conferencing services requires a systems of servers, and this would be a very expensive undertaking. However it should be possible in many cases to build on services (such as Skype) that already exist.
The trick here is to make use of the many commercially available services and come up with an integrated solution.
One possible scenario where you may consider a dedicated service is for customer support - in this case the users are communicating with a central support desk in a hub-and-spoke topology (as opposed to an any-to-any topology which makes life much more complicated) The support person would be able to talk to the customer using audio, or even video.
Cost and technical implications: Difficult to say - this needs to be considered very much on a case by case basis. Cost effective solutions should be possible using existing providers.
The customer support example would also be in the medium to high cost range, both in terms of implementation and on-going running costs.