You want to make animation in which UFO comes and flies between the buildings and scanns the environment using lasers. You take the camera and go outside to record the footage , which you use in your animation. You come home and start ripping footage mask the environment of the recording so that the modeled UFO seemed to move between buildings and the scanning laser light lingered on the buildings …
After an hour, you have a short animation .
Now imagine that you have connected a camera to the computer and the camera is connected with Blender. There is created a script that continuously detects the buildings and roads from the footage (like the Face Detection function in cameras ) . Automatically creates masks in Blender pre-selected elements - buildings or roads … Camera image has a depth which is recognized by Blender .
Your UFO model has entered the information to navigate the (over) detected in the image paths (roads) and spontaneously turn on and off the lasers.
Wherever you point your camera UFO will follow the direction of passing buildings or hovering over them if it will be unable to avoid it …
Ohter example:
You have set few cameras in your home and got few monitors. When you enter the home on monitor will appear a image from the camera with modeled by you person -you virtual friend that will follow you in the picture or do what you told it to do (earlier programmed orders)
This may work as a virtual guide for people for example on airports etc …
That’s just an example of what in future may appear in Blender. It’s not just an “closed” idea but a vision of a functions that can appear in the future with lots of possibilities.
Opensource idea for public or private use with no-coplicated interface because all (or partially) would be done by programmers …
If someone thinks this is stupid than sorry for your time !
No one is forbidden to dream …