Difference between revisions of "Video-based AR"
| (20 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
| − | ''' | + | '''Camera passthrough''' is a capability of a headset to pass through a video feed to the user on the headset's display. It shows the outside world to the user so he does not become disoriented. It is also known as "video passthrough". It can be used in a [[mixed-reality headset]]. |
| − | + | The idea of video passthrough was popularized by the [[Lynx R1]] and further popularized by the [[Quest 2]] and [[Quest Pro]]. | |
| − | One of the difficulties of | + | One of the difficulties of doing passthrough right is the positional difference between cameras on the outside of a headset and the eyes of the user. So the image has to go through a distortion process. On the Quest Pro and [[Quest 3]], this uses a neural network. Other gadgets potentially also use a neural network. |
| + | |||
| + | It is possible to use a [[light field camera]] to capture an image and reproject it at a different transform. | ||
| + | |||
| + | It is also possible to have a camera offset from the headset and facing the headset, and put a mirror on the headset. | ||
| + | |||
| + | ==See also== | ||
| + | * [[See-through AR]] | ||
| + | |||
| + | ==References== | ||
| + | {{Reflist}} | ||
| + | |||
| + | [[Category:Technologies]] | ||
| + | [[Category:MR headset]] | ||
Latest revision as of 23:50, 31 March 2025
Camera passthrough is a capability of a headset to pass through a video feed to the user on the headset's display. It shows the outside world to the user so he does not become disoriented. It is also known as "video passthrough". It can be used in a mixed-reality headset.
The idea of video passthrough was popularized by the Lynx R1 and further popularized by the Quest 2 and Quest Pro.
One of the difficulties of doing passthrough right is the positional difference between cameras on the outside of a headset and the eyes of the user. So the image has to go through a distortion process. On the Quest Pro and Quest 3, this uses a neural network. Other gadgets potentially also use a neural network.
It is possible to use a light field camera to capture an image and reproject it at a different transform.
It is also possible to have a camera offset from the headset and facing the headset, and put a mirror on the headset.